mobile os overview april 2011

North America United States TMT Wireless Equipment

11 April 2011

Mobile Operating Systems Choices are multiplying

Jonathan Goldberg, CFA Research Analyst (+1) 415 617-4259 [email protected]

Brian Modoff Research Analyst (+1) 415 617-4237 [email protected]

Kip Clifton Research Associate (+1) 415 617-4247 [email protected]

Deutsche Bank Securities Inc.

All prices are those current at the end of the previous trading session unless otherwise indicated. Prices are sourced from local exchanges via Reuters, Bloomberg and other vendors. Data is sourced from Deutsche Bank and subject companies. Deutsche Bank does and seeks to do business with companies covered in its research reports. Thus, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MICA(P) 007/05/2010

FITT Research

Fundamental, Industry, Thematic, Thought Leading Deutsche Bank’s Company Research Product Committee has deemed this work F.I.T.T. for our clients seeking a differentiated view on the smartphone market. Our wireless technology team looks at long-term trends shaping the mobile operating system landscape which is emerging as a key battleground for the future not only for mobile but for all computing.

Fundamental: The landscape is changing rapidly, making a long term impact Industry: Android has numerical superiority but will not become ‘dominant’ Thematic: Developers matter most, and Apple offers them the best returns Thought-leading: The Web will win, and the fight is on to shape HTML5

Co

mp

any

Glo

bal

Mar

kets

Res

earc

h

North America United States TMT Wireless Equipment

11 April 2011

Mobile Operating Systems Choices are multiplying Jonathan Goldberg, CFA Research Analyst (+1) 415 617-4259 [email protected]

Brian Modoff Research Analyst (+1) 415 617-4237 [email protected]

Kip Clifton Research Associate (+1) 415 617-4247 [email protected]

Fundamental, Industry, Thematic, Thought Leading Deutsche Bank’s Company Research Product Committee has deemed this work F.I.T.T. for our clients seeking a differentiated view on the smartphone market. Our wireless technology team looks at long-term trends shaping the mobile operating system landscape which is emerging as a key battleground for the future not only for mobile but for all computing.


All prices are those current at the end of the previous trading session unless otherwise indicated. Prices are sourced from local exchanges via Reuters, Bloomberg and other vendors. Data is sourced from Deutsche Bank and subject companies. Deutsche Bank does and seeks to do business with companies covered in its research reports. Thus, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MICA(P) 007/05/2010

FITT Research

Fundamental: The landscape is changing rapidly, making a long term impact We expect a continued pace of weekly changes in the state of the market this year, but decisions being made today will eventually lock in future performance. Smartphone growth is strong and we think it will accelerate even further as prices for components fall. We forecast smartphone share into 2015.

Industry: Android has numerical superiority but will not become ‘dominant’ Android has taken the lead in shipments and could attain leadership in installed base in a few years. Despite this numerical superiority, we think flaws in the platform (like fragmentation) mean that Google will not become the dominant vendor. The smartphone OS landscape will not resemble PC landscape. This leaves open the door for other vendors to potentially carve out profitable or strategic niches, which leaves room for Microsoft to pick up the pieces if Google stumbles.

Thematic: Developers matter most, and Apple offers them the best returns Hardware is a commodity, and the only way for phones to differentiate is through software. This puts priority on building app developer ecosystem. Android’s installed base still lags Apple’s. More importantly, by our math, the revenue opportunity on Apple is an order of magnitude higher than on Android. We calculate hypothetical returns and find the frictions in the Android Marketplace are so high that monetizing apps is difficult. This is a major hurdle, and has opened the door to others, like Amazon, which risks further fragmenting the developer base.

Thought-leading: The Web will win, and the fight is on to shape HTML5 Over the long term, we believe the only force that can bridge the splintered landscape will be the Web itself. Most OS vendors seem to share this view. We expect a new round of browser wars, but this time around the battle will take place within the bodies that define web standards, particularly the nascent HTML5 standard. Obscure debates over topics like video codecs will play an important role in shaping the future landscape. We also see a far-ranging change in the Web itself and the way in which we interact with the Internet.

Investment options: Component vendors to prosper, giants are challenged In our view the great irony of this rapidly growing market is that there is no easy way to invest in a mobile operating system. Smartphones run either proprietary OS, captive within larger companies or provided free by Google. Nonetheless, smartphones are now computers in their own right, and consumers will continue to find new ways to use them. We believe this will drive the need for more advanced radios, basebands and processors. It will also likely boost demand for cameras, speakers and sensors. We believe there are still many component vendors who can benefit from this, a factor which may not yet be reflected in their share prices.

11 April 2011 Wireless Equipment Mobile Operating Systems

Page 2 Deutsche Bank Securities Inc.

Table of Contents

Executive summary ........................................................................... 4 Expectations ...................................................................................... 6 Developer math ................................................................................. 7 State of the mobile OS.................................................................... 10 Smartphone growth........................................................................ 11 The nature of the Web ............................................................................................................ 15 Move toward the cloud........................................................................................................... 20 OS landscape ................................................................................... 23 Installed base .......................................................................................................................... 25 App Store update.................................................................................................................... 27 Security ................................................................................................................................... 29 Cheat sheet ...................................................................................... 30 How to challenge a giant ........................................................................................................ 30 Individual OS review....................................................................... 31 Android............................................................................................. 32 Fragmentation......................................................................................................................... 34 Conclusion .............................................................................................................................. 35 Summary................................................................................................................................. 36 Android – Other than Google ......................................................... 37 These are not the Androids you were looking for ................................................................... 37 Symbian/MeeGo.............................................................................. 38 End of an era........................................................................................................................... 38 RIM.................................................................................................... 40 RIM is fragmenting itself......................................................................................................... 40 HP webOS ........................................................................................ 43 Surprise underdog or just an underachiever?.......................................................................... 43 Microsoft Windows Phone ............................................................. 45 The best defense is a good offense........................................................................................ 45 Apple................................................................................................. 49 It is good to be in the lead ...................................................................................................... 49 Adobe Flash ..................................................................................... 51 Beneficiaries..................................................................................... 52 Faster, higher, denser ............................................................................................................. 52 Components ........................................................................................................................... 55 Future features ................................................................................ 59 Feature phones ....................................................................................................................... 59 Mobile payments .................................................................................................................... 59 Augmented Reality.................................................................................................................. 60 M2M – Skynet’s First Steps.................................................................................................... 61 The nature of phones.............................................................................................................. 62 Appendix A: Cellular standards roadmap ..................................... 63


Deutsche Bank Securities Inc. Page 3

Table of Figures Figure 1: Smartphone landscape summary...............................................................................4 Figure 2: Smartphone OS market share....................................................................................4 Figure 3: Hypothetical developer revenue ................................................................................8 Figure 4: Hypothetical app market sizing by platform...............................................................8 Figure 5: Android ecosystem sensitivity to conversation rates and installed base ...................9 Figure 6: Android ecosystem sensitivity to app downloads and installed base ........................9 Figure 7: Windows ecosystem sensitivity to app downloads and installed base .....................9 Figure 8: Smartphone landscape summary............................................................................. 12 Figure 9: Smartphone OS Market Share ................................................................................. 23 Figure 10: Our view on future OS market share ..................................................................... 24 Figure 11: Our longer term view of smartphone market share ............................................... 24 Figure 12: Installed base by smartphone OS .......................................................................... 26 Figure 13: Installed base by smartphone OS .......................................................................... 26 Figure 14: Application Store – applications available as of latest date .................................... 27 Figure 15: App Store growth, from launch to one year, number of apps................................ 28 Figure 16: App Store growth, from launch to 2.5 years, number of apps............................... 28 Figure 17: Android installed base by OS version..................................................................... 34 Figure 18: Android versions by OS release ............................................................................. 35 Figure 19: Nokia’s mobile OS forecast.................................................................................... 38 Figure 20: webOS developer activity as measured by RSS feeds of the App Catalog ........... 44 Figure 21: Growth of all iOS apps ........................................................................................... 49 Figure 22: Number of iPad vs. iPhone apps, days after launch of App Store.......................... 50 Figure 23: Median mobile phone clock rate (MHz) ................................................................. 53 Figure 24: Median mobile phone ROM capacity (MB) ............................................................ 54 Figure 25: Median mobile phone RAM capacity (MB) ............................................................ 54 Figure 26: Cellular standards roadmap.................................................................................... 63



Executive summary In this report we look at the developments in the mobile operating system (OS) landscape. Many people believe the smartphone market would look like the PC market 30 years ago with one OS eventually becoming dominant. In this comparison, they substitute Android for Windows, and Apple for Apple. We believe this view is mistaken. Instead, we think the landscape will remain fragmented for many years. The only platform which we think can unify the landscape is HTML5 and web tools, and these are still several years from maturity.

Figure 1: Smartphone landscape summary Vendor 2010E units (m) Installed base (m) App Store Apps New models shipped

in 2010 Tablet models in 2011?

Symbian 100 ~300m 30,000 10 0

Windows Phone < 2m ~20m (incl. WM) 11,500 3 0

Android >50m ~50m >200,000 >100 >100

Limo/Lisp ~10m 100m ? ? 0

iOS ~100m ~100m >300,000 1 3?

RIM 50m 60m 10,000 5 2?

Palm <2m 5m 5,056 0 1

MeeGo 0 0 0 0 0

Bada <5 m <5m ? 3 0Source: Deutsche Bank, company data

In this kind of fragmented landscape, many vendors will be able to carve out niches based either on some specific demographic or along individual strategic lines. Android unit volumes are already the leading smartphone OS, but we see no one who, as a result of this, wants to short Apple. Instead, we think Apple will continue to grow its very profitable ‘niche’. Android will grow faster, but the vendors making these devices will struggle with profitability levels amid a sea of undifferentiated phones. RIM’s QNX, HP’s webOS and Microsoft’s Windows Phone 7 all have the potential to build viable niches as well. Moreover, we think there will be new entrants to the field as various market participants look to steal a march on future trends and build HTML5-based alternatives or twist Android to their own ends.

Figure 2: Smartphone OS market share

RIMM15%

PALM/webOS2%

Apple19%

Windows Mobile

10%

Symbian & MeeGo

45%

Android1%

Linux/Other8%

2008

RIMM17% PALM/webOS

1%

Apple27%

Windows Mobile

0%

Symbian & MeeGo

33%

Android17%

Linux/Other5%

2010E

Source: Deutsche Bank and company data

A key component of this fight will be winning the loyalty of app developers which help each platform differentiate. For developers, Apple’s iOS remains the focus. Android has become the default second choice, but developers continue to struggle monetizing this platform. We see two important elements to this analysis. The first is installed base. Despite the growing



market share of Android phones, the installed base of all iOS devices remains far larger. Market share estimates tend to only look at smartphones, but we think it is important to count iPod Touch and iPad units as these constitute a large part of the addressable market for app makers. As such, we think the iOS installed base was about 121 million at end-2010, while the Android installed base was around 50 million. The Symbian installed base is larger, but has been rendered effectively zero by Nokia’s announced intention to move away from that platform. We think the Android installed base will catch up with Apple this year. However, our second tenet is that developers will still prefer iOS because making money on Android is difficult for most developers. By our math, the revenue opportunity for developers is an order of magnitude higher on iOS than on Android. We detail this math below.

One of the great difficulties in reviewing this space is the lack of suitable investment options to tap into this growth. We think there is little correlation between Google’s or Microsoft’s share price and the growth of Android or Windows Phone 7 this year. Aside from buying Apple, there are a few direct investments in the growth of smartphones and the maturity of mobile OS. However, we think there are several handset suppliers who stand to benefit. Smartphones are becoming more powerful, and they are being used for a wider array of tasks. This means handset vendors will need to cram more components and features into their devices to tap into consumer demand. This should prove to be good news for a variety of suppliers. We would include: the processor companies (QCOM, ARM, and CEVA); the radio components (Skyworks, Avago); audio component vendors (AAC Acoustic, Wolfson, etc); camera and video makers; touch screen controllers; connectivity vendors (e.g. BRCM, NXPI); and arguably, memory vendors. In this note, we analyze trends around the growing density of phones, which has nearly doubled in the past ten years. We expect this trend to continue. We also discuss ancillary beneficiaries in areas like mobile payments, augmented reality and M2M.



Expectations Writing a report like this is something like running on a treadmill. In the time between pen-to-paper and publishing a lot can change. We began this report in January 2011. In the interim, we have had Nokia’s choice of Microsoft, HP’s re-launch of webOS, RIM’s Playbook launch, several iterations of revisions to Android’s openness and any number of new product launches.

Sometimes, this makes it hard to see the forest for the trees. To counter that minute-by-minute view of the details of the landscape we want to present a few changes we expect to see to the landscape this year. We see these as separate from our higher level investment thesis on the subject which we outline in the reports. We offer these as tactical suggestions for what we think may happen.

Firstly, Android is likely to continue to grow rapidly. Advances in silicon mean that price points for sufficiently powerful smartphones are reaching mass market levels, and most of these phones will run Android. We think every headline for some time from the Street – third party data firms and the blogosphere– will point to some other way in which Android share is advancing.

Secondly, we expect to hear more about legal issues relating to Android. Oracle is already suing Google for Android patent issues. Many other members of the Android community are also involved in similar litigation. While we have no idea how any of these law suits will turn out or if others may emerge, we think this is a topic that a few people are discussing. Over the next year, we think this reality will seep into the industry’s consciousness and gain further attention.

Microsoft will remain persistent. We think sales of Windows Phone 7 will remain modest. We estimate that a full slate of Nokia WP7 products will not be on the market for another year, and the other WP7 vendors appear to be losing interest. Despite that, Microsoft has the resources (both financial and energy) to stay in the game. We think they will continue to aggressively market the OS and encourage carriers to support it.

Finally, we think several new operating systems will enter the market. Some of these will be Android variants from other web properties. Others will be HTML5-based stepping stones towards what we consider to be the future of mobile, web tools-based devices. Others will not be truly new OS, but instead platforms for feature phones that emulate many of the key features of smartphones such as social networking integration. The press has already reported that Baidu, Facebook and Motorola are all working on variants of these. We think these reports have some truth to them, and believe others will emerge with time.



Developer math In this note we will tend to focus on each OS from the point of view of developers, and try to gauge each OS’s appeal on that basis. While we cannot exactly quantify that appeal, we do offer the formula below as the framework for our discussion. This will not provide a numeric answer to the question, but we present it by way of introducing the assumptions on which we base our analysis. Put simply, each OS holds appeal for developers as a function of:

Installed Base * Marketability * App Value-- Cost of Development

The installed base is the number of users who own a device that runs each OS. We provide estimates of installed base for each platform.

Marketability – this is the number of users that could potentially end up buying an app. This is a function of user behavior, how many people actually use apps on each platform. It is also a function of the friction involved in making a transaction on each platform. These are both highly qualitative assessment and impossible to calculate accurately.

App value – refers to either the price of a download or the lifetime value of having a user with that app. This includes services subscriptions, ad revenues, premium upgrades and other in-app purchases. This then needs to be adjusted for revenue share with each platform provider. We do not have the data to calculate these numbers today, but believe that with time we could eventually put some more solid math behind this calculation.

Cost of development – this includes resources needed to write the app and is largely measured in engineering person hours.

We explore each of these elements in general in the following sections then in detail in section on each individual OS.

We think looking at it from this perspective adds an important dimension to the discussion of OS. Today, most comparisons among OS look at the number of apps. We think this is meaningless as not all apps are created equal. More important for developers is to calculate how much money they can make off each platform. Dealing with Apple is not easy, but the installed base is large, most users download apps and the App Store works very well. There are also numerous ways to monetize each app. By contrast, Android has a rapidly growing installed base and is easy to develop for, but suffers on the other two metrics. Emerging OS like Windows Phone and webOS have a relatively low cost of development but are playing to almost no installed base.

Let’s put some numbers behind this.

First, to be clear, there is no solid data on this topic. We have seen dozens of individual developers’ blogs about the performance of their apps on various platforms. There are also about a dozen analytics companies with published data on the topic. So far, we have seen nothing truly comprehensive. Apple, Google and Microsoft have this data. We believe all smartphones based on these operating systems regularly report back usage data to their respective motherships. However, none of them have shared this data, in part, we believe because they do not want to draw attention to this particular reporting feature.

For example, let’s say you are a developer. You have developed a game. On the desktop, you would sell this game for anywhere between $20 and $40, but that still requires a publisher, physical hardware and distribution costs associated with selling to Best Buy. Apps on the Apple App store sell for a large array of prices. One of the top grossing, Zombies Versus



Plants, sells for between $0.99 and $4.99 on the iPhone and typically $9.99 on the iPad. There is a significant body of bloggage about developers experimenting with prices, promotional price cuts and freemium models. We think standard practice on the App Store is to give away a ‘Lite’ or Trial version of the app for free or $0.99 and separately sell a $5 fully-featured app that includes more gaming levels or additional characters. Microsoft has actually built a Trial mode into the OS. The top grossing app recently is Angry Birds which is $0.99 on iTunes and free on Android. For this example, let’s say the Lite version of the app sells for $0.99, which is probably the median app price on iOS, and the full version of the app is $4.99

The next question is of downloads. A large problem on all these platforms is ‘discoverability’, that is the ability to make consumers aware that an app exists. This is a thorny problem, with many companies including Chomp emerging to address. Currently, the best way to gain awareness is to be prominently featured on the Apps store. The process for achieving such status is completely obscure, and we sense it is something that even Apple does not fully manage. A top ten or even a top 50 apps, does significantly better than anything further down. We refer to 2D Boy’s blog (2dboy.com) which provides us with some interesting insights on the trends for one app. When they launched the iPad version of their app, they were featured on the App Store and saw 125,000 downloads during the first month. Without prominent featuring, downloads are at best half of that level.

The next question would be what percent of users upgrade to the premium version. Our sense is that this rate is in the low single digits. We also think conversion is much higher on iOS and Windows than it is on Android. We argue below that payment friction on Android is very high due to a variety of factors. Based on these assumptions, developer revenue works out to something like the data in Figure 3. This would seem to indicate that even without a prominent App Store placement, the revenue opportunity is much higher on Apple. An important difference is the fact that Android users are very reluctant to purchase apps on the marketplace and as a result the majority of Android apps are free. If consumers were willing to pay $0.99 for an app, the situation would be drastically different and the revenue would be close to iOS. In the formula above, this is what we refer to as marketability, and it is an area where Android struggles.

Figure 3: Hypothetical developer revenue Downloads Lite Price Conversion Rate Full Price Total Revenue

iOS 50,000 $0.99 3% 4.99 $56,985

iOS Featured 100,000 $0.99 3% 4.99 $113,970

Android 50,000 $0.00 3% 4.99 $7,485

Windows Phone 1,000 $0.99 3% 4.99 $1,140Source: Deutsche Bank

We also examined this from a top-down perspective wherein we look at the installed base and the number of apps each user, on average, downloads per month. This gives us a rough sense of the size of the whole market, but we emphasize that it is very rough because of the assumptions around pricing. The results (Figure 4) are similar.

Figure 4: Hypothetical app market sizing by platform Installed

BaseApps Per

Phone Per Month

Total Downloads

Lite Price Full Price Conversion Rate

Total Revenue

iOS 121 10 1,208 0.99 4.99 3% 1,377

Android 52 10 520 0.00 4.99 3% 78

Windows Phone 1 12 12 0.99 4.99 3% 14Source: Deutsche Bank



To better gauge the sense of the problem we also tested the Android developer revenue opportunity’s sensitivity to a variety of factors. First, we looked at a sensitivity to conversion rate, then we looked at the number of apps per month, and finally compared both of these to the size of the addressable market. By this math, the most important factor is the size of the installed base. All else being equal, the Android developer opportunity cannot match the iOS opportunity until the Android installed base reaches 800 million, from roughly 50 million at the end of last year. We think it can reach that size, but not until 2015. If Android users started doubling the number of apps they download, then could pull that date forward by a year, but that seems difficult to achieve.

Figure 5: Android ecosystem sensitivity to conversation rates and installed base Conversion Rate

78 1% 2% 3% 4% 5%50 25 50 75 100 125

Installed 171 85 171 256 341 427Base (m) 250 125 250 374 499 624

400 200 399 599 798 998500 250 499 749 998 1,248800 399 798 1,198 1,597 1,996

Source: Deutsche Bank

Figure 6: Android ecosystem sensitivity to app downloads and installed base Apps Per Month

78 5 10 15 20 2550 37 75 112 150 187

Installed 171 128 256 384 512 640Base (m) 250 187 374 561 749 936

400 299 599 898 1,198 1,497500 374 749 1,123 1,497 1,871800 599 1,198 1,796 2,395 2,994


We then compared these figures to the Windows opportunity. WP7 has a tiny installed base today so the opportunity today is essentially immaterial. However, their willingness to share some of this data points to a better ‘Marketability’ on this platform. This is extrapolating considerably, but shows how much easier it is for this platform to become profitable than Android with a much smaller installed base.

Figure 7: Windows ecosystem sensitivity to app downloads and installed base Apps Per Month

14 5 10 15 20 2550 285 570 855 1,140 1,425

Installed 171 974 1,949 2,923 3,898 4,872Base (m) 250 1,425 2,849 4,274 5,699 7,123

400 2,279 4,559 6,838 9,118 11,397500 2,849 5,699 8,548 11,397 14,246800 4,559 9,118 13,676 18,235 22,794




State of the mobile OS Two years ago we published a FITT report looking at the mobile landscape titled Anarchy in the OS. As the title implies, we noted a huge degree of fragmentation across the mobile OS landscape. Since then much has changed, but we are no closer to a common OS platform for mobile phones than we were two years ago. In an update piece that we published in March last year, Digits #20 The Big Idea, we updated our thesis on the space with the idea that over time the only unifying force for the mobile OS will be the Web itself and the use of web tools for developers. Slowly, the industry is moving in this direction.

We think a few trends have become apparent:

Smartphones are exploding, and will continue to grow rapidly.

Android has tremendous momentum. It already has the largest share of smartphone sales, overtaking Symbian in December. And this is only the beginning. We think its growth will accelerate this year and again in 2012 as smartphone component prices continue to fall sharply.

Despite its success, Android has many flaws. These are unlikely to put a dent in its momentum but will likely leave open the door for others to carve out niche markets or grow their already highly profitable segments.

With even Google talking openly about the eventual move towards web-based operating systems, we expect the fight this year to center on the features of HTML5, the next generation of the web standard towards which many OS makers aspire. Key among these will be the codecs for video and the <video> tag, cloud synchronization, and the struggle between apps and web pages as metaphor.

Developers will go where the money is. Android offers volume but uncertain monetization. Apple’s iOS remains far more lucrative for app developers. All other platforms remain works in progress.

In short, the industry remains in flux, and we expect many companies to continue to find ways to build strategic advantage through various approaches to mobile operating systems.



Smartphone growth One of the clearest trends in technology today is the growth of smartphones. Our current handset forecast estimates 39% growth in 2010. We are still compiling end-of-year numbers and suspect that estimate will prove to be too low. Similarly, we think growth in 2011 will exceed our 28% target. Most importantly, as we pointed out in our sister publication Signals to Noise last week, we believe smartphone growth will accelerate in 2013 as powerful new processors enter the market at far lower price points. In particular, we believe that by 2013 we will see volume shipments of 1 GHz processors in phones priced at USD100. At that point, feature phones will quickly disappear, with survival dependent on the pace of consumer education and carrier data plan pricing.

We think smartphone growth will have a profound impact on broad technology trends along two axes. First, as noted above, smartphones are now essentially as powerful as laptops. Most of the population of the world will never own a PC, but within a few years most of them will own a smartphone. The phone will become the data device for most people. Second, the kind of things that people use their phones for will greatly expand. By comparison, the changes brought upon by the advent of the Internet a decade ago will seem minor in comparison. In developed markets, we expect to see a significant change in the way that people shop and interact. In emerging markets, we are already getting a glimpse of the impact of having unfettered access to data on the political process. Not all the effects will be positive. For example, those concerned that video games dumb down children, will likely rue the effect of having video games become completely portable Ask any parent with an iPhone. This is not the venue to expand on the wonders of technology; we will save that for Wired and the tech press. Suffice it to say that the Mobile Web will be a powerful force.

Big picture thesis We believe that smartphone hardware is rapidly become a commodity. By this, we mean that any hardware feature can quickly be replicated, providing no sustainable barrier to entry. There is still immense room for improvement and we expect to see some very impressive improvements in industrial design in coming years. However, all of this can be copied. In some cases, we have seen top tier handset models copied before the phone even reached the market.

In this kind of environment, the only way to differentiate a phone is through improved software, brand building and distribution. These are all tricky and often rest outside the expertise of many incumbent vendors. The obvious example of this sort of differentiation is Apple’s iPhone which has a powerful combination of software expertise, marketing prowess and a wide retail network to distribute its products. No other company can match these precisely.

By software, we are referring to two elements: first, the operating system (OS) of the phone including its user interface (UI) or user experience (UX) and second, the ecosystem of third-party software applications available for the phone. These concepts are clearly linked. And for a phone or a company to be successful both elements need to be in place.

Operating system landscape In our first report on the topic two years ago, we identified seven prominent smartphone operating systems. There are nine today, but only a few of these are shipping in any true volume today. We provide of a summary of the nine in Figure 8 and detail the outlook for them in a later section.



Figure 8: Smartphone landscape summary Vendor 2010E Units (m) Installed base (m) App Store Apps New models shipped

in 2010 Tablet models in 2011?

Symbian 100 ~300m 30,000 10 0

Windows Phone < 2m ~20m (incl. WM) 11,500 3 0

Android >50m ~50m >200,000 >100 >100

Limo/Lisp ~10m 100m ? ? 0

iOS ~100m ~100m >300,000 1 3?

RIM 50m 60m 10,000 5 2?

Palm <2m 5m 5,056 0 1

MeeGo 0 0 0 0 0

Bada <5 m <5m ? 3 0Source: Deutsche Bank, company data

Changes in the OS landscape In recent weeks, it has become clear that Android has overtaken all other OS unit shipments. Despite this, the landscape remains highly fragmented. The total cell phone market in 2010 was about 1.4 billion units, of which around 20% or close to 300 million were smartphones. When viewed on this basis, it should be clear that the landscape is still fragmented. While Android led the market in December, for the year it only had 20% of the smartphone market, a small piece of the overall market.

We think this fragmentation is important. Despite Android’s success, we think there is still room for other platforms to exist. There are also powerful forces that will likely mean the market remains fragmented for many years to come. On the other hand, we think the economics of software will pressure the market to move towards a common platform that works on all phones. Our view, which we detailed in Digits #20 (The Big Idea, March 2010), holds that the only solution to this dilemma will be web tools. The only force capable of reconciling the industries players and providing a unified platform will be the web.

The idea of web tools is itself very complicated but we expect it to be the key battleground in coming years. We had originally viewed this trend as meaning that eventually cell phones (and probably all computers) would eventually become very sophisticated browsers. It turns out the vision is more complicated than that.

So first let’s look at what these devices will have in common. We believe that ultimately smartphone run times will be entirely accessible by common web programming tools using HTML, CSS and JavaScript. To simplify, developers who program apps today using advanced languages will eventually be able to use far simpler set of programming languages. This will open up app development to a far wider range of developers. There are probably a few million programmers today working in C, C++, Objective C and the like. By contrast there are hundreds of millions of people who can write in HTML. Most high school students in the US today learn basic HTML skills, if for no other reason than they learned how to customize their MySpace pages. This will also open the market to websites that even further simplify the design process. For instance, as MySpace wanes many teenagers have started to create custom sites using Tumblr, a popular blogging app that is actually a disguised HTML authoring tool. As smartphones move to HTML environments we expect an explosion of development.

In many areas, these tools are already accessible but are yet to fully move into mobile. This delay is in part due to the early stage of the smartphone industry. One area of investment we think could prove particularly interesting for the venture community would be in this area. The search for the mobile Tumblr is on.



However, there are many areas in which web tools are not yet ready for mobile. HTML was originally designed for desktop PCs with always-on connections to the Internet. For the standard to proliferate on mobile, it needs to be able to operate in areas where there is no network connection. The standards body behind HTML, the W3C, has been working on this problem. The next version of the standard, HTML5, is being designed to make better use of a device’s local capabilities. That is to say browsers of the future will be able to execute code without an Internet connection. They will be able to use the local processor and be able to store large amounts of data on the device.

Again, many of these features are available today, and are slowly being incorporated into devices and web pages. But the standard is not complete, and work still needs to be done on a few important areas. The most contested of these, so far, is the video component. HTML5 includes a new function for video, <video>, but much of the specs for this component are still being ironed out. Among the difficulties here are agreeing to a common video codec or compression scheme. Codecs are important for ensuring minimum bandwidth requirements for video streams. The most common codec today is H.264 and is commonly used on most video web sites. Most of the HTML constituents are comfortable with this standard which has been around for many years. The issue is the associated intellectual property (IP) and the patents on which the standard is based. These are owned by MPLA, a media industry patent pool. MPLA requires content providers, the companies that stream video, sign a license for use of this technology. Some in the web community are concerned that MPLA will eventually begin charging for the license. Google, concerned about this, launched an alternative codec called WebM which is entirely open-sourced. MPLA responded by extending the current license until 2015 and then further extending it saying H.264 will be royalty free for its lifetime. This offer is actually not as straightforward as it sounds, but conversely no one has tested WebM’s patents so there may be patent holders there who could emerge. The debate about web video is ongoing, and it is too soon to call a winner.

To further complicate matters the existing de facto web standard for video is not an open license at all; it is Flash platform. This is an important revenue stream for Adobe, or more precisely the tools used to create Flash video. This is the root of a public debate between Apple and Adobe. Our view is that Flash will likely remain a video platform for the mobile web, but will gradually give way to the common HTML <video> format. On the PC web, Flash is not only used for video. It is a complete programming environment in its own right and is used to write ‘apps’ for web pages today, especially games. Adobe had hoped to make this a common platform for mobile apps as well, in effect a competitor to HTML5. This now seems unlikely to occur. Aside from Apple’s refusal to implement Flash on the iPhone, there are technical barriers that make this unlikely. Getting Flash to work on different processors and different operating systems remains a surprisingly cumbersome project. Essentially, Flash has to be ported to each version of a chipset. This is immensely time-consuming and labor-intensive. While Flash (may) have some performance advantages over HTML5, the time-to-market issue will likely prove to be a significant barrier to widespread adoption.

It should be clear that the subject of web video will not be resolved soon and thus HTML5 will take some time to achieve its goals. And this is just one aspect of HTML5 being resolved. At heart, the key difficulty faced by the World Wide Web Consortium (W3C) and HTML5 is the same issue that has plagued the mobile community for years. Smartphones are incredibly diverse with innumerable combinations of processors, screen sizes, form factors and input methods. HTML5’s task is to unify all of these in a common framework, but device difference will persist.

Until HTML5 becomes more mature, OS makers will continue to deploy various tools to bridge this gap. As a result, we have seen many of them launch alternative software environments. One of the key elements of an OS is the Software Development Kit (SDK) that



each OS provides to developers. SDKs have three key elements that rest at the heart of a developer’s ability to tap into a particular OS. These three are the Application Programming Interface (API), an emulator and developer tools. (SDKs contain many more elements, but these are the three major ones).

APIs are the list of functions or commands available to programmers, and each SDK contains a dictionary-like list of APIs that describe all the ways in which they can be used. To put this in context, version 4 of Apple’s iOS contained 1,500 new APIs to the thousands available in previous versions. APIs are a key metric for developers, measured in quality and not only quantity. Google, for instance, withholds APIs of each new version of Android for several months, giving one hardware partner an advanced set to give them a short-term advantage in the market for launch of flagship devices. There is also a debate about so-called “private” APIs for use only by the OS designer to build special apps. In iOS, the various multi-task functions are controlled by Apple. In Android, Google restricts access to Google apps like Gmail and Google Maps. (This is a complex issue which we will save for a further note.)

Emulators are tools that run on desktop computers which can simulate how an app will perform on a phone. This is useful for testing new apps and is emerging as an important differentiator for developers as well. There is a perception that Android’s emulator offers a better likeness of actual app performance than Apple’s, but as with all technology there are trade-offs and no perfect solution.

Finally, each OS comes with a set of developer tools that perform a variety of functions. These range from Apple’s well-regarded X-Code authoring tool (essentially a very advanced word processor for typing code) to Nokia’s Qt toolkit which aids in developing common UIs across multiple platforms (and potentially offers much more, please see Digits #25 for more on this).

So far this is all relatively straightforward. However, each OS comes with its own set of design principles. Those are expressed in the SDK, but do not always live up to the reality of developing in the real world. Many device makers have found that their original intentions cannot fully address the device diversity and have begun to add a more advanced set of APIs. The most important of these currently is Android’s Native Development Kit (NDK). HP Palm’s webOS was originally built to be entirely run by web tools, but due to the shortcomings in HTML, they eventually released a Physical Development Kit (PDK). RIM’s new QNX is also built with web tools in mind, but also offers Adobe AIR to let developers write more advanced apps. Even the iPhone began life with only web tools available for app development before releasing the full iOS with the launch of iPhone 3G.

This is a blur of names, but at heart, these alternative SDKs perform the same function. They let developers take advantage of lower level processor functions. The key property here is to tap into hardware acceleration for graphics and audio processing. Despite the appeal of web tools, they are ‘higher layer’ or ‘more abstract programming languages. The goal here is to abstract or decouple the app software from the underlying hardware so that devices can bridge multiple hardware platforms. Apple is the furthest along here because they only ask developers to support two highly similar devices (iPad and iPhone/Touch). Higher level languages essentially insert a level of translation in between the app and the hardware. This is highly desirable but can result in either a performance penalty or a lack of access to device functions, and sometimes both.

We see all of these as necessary stop-gap solutions that could persist for a number of years. Like the “Temporary” buildings built during World War II which housed classrooms at UC Berkeley into the 1990s, we think these patches will persist until these issues are solved by HTML5.



This problem looks so complex that it may never be solved and devices could potentially always carry an element of proprietary programming. Critics of HTML5 point out that new features are continually being added to smartphones and the ponderous pace of deliberation of the W3C means HTML will always be several steps behind the curve. We think time will eventually heal this issue, but several elements will need to be added to the standard and the timeline is not entirely clear for their inclusion. They include:

Hardware graphics acceleration. As noted above, the ability for an app to tap into extra processing power for graphics is important, especially for game developers. It will be a long time before all phones offer this, so each OS and HTML will have to take this into account.

GPS. The W3C is working on a specification for Geolocation but we believe this still requires significant work, particularly around building common support across multiple location data sources (GPS, location databases, etc.)

Device sensors. One of the key hardware advances of the iPhone was the addition of motion sensors. We imagine that with time devices will have other similar sensors to detect real-world phenomenon like temperature and light. This remains a work in progress for web standards.

Connectivity. Wi-Fi and especially Bluetooth have a wide range of software profiles that need to be supported. To achieve a truly web-only device this will need better support in HTML.

Screen size. Perhaps the trickiest issue in smartphones will be the diversity of screen sizes and resolution. This problem has largely been solved on the PC-based web, but those computers have the advantage of ample processing power and screen real estate. The issue will prove particularly troublesome in mobile for graphics intensive apps such as games and design-heavy web sites.

The nature of the Web

Our belief is that with time all of the abovementioned issues can be resolved. The web community may not be able to find a perfect solution, but should be able to find one that works well enough for most. However, we think there remains another emerging divide in the mobile web. This is a more existential question of what does having the web or the Internet on a device truly mean. This sounds somewhat abstract but will likely become a key issue in years ahead.

Over the past decade of the Internet’s existence we have grown accustomed to accessing it through a common medium - the web browser. Users open a browser, the browser sends a request to a web server which responds with various pieces of data. The browser then renders this data into a readable format on the user’s computer screen. Users then navigate to various other web pages and the process repeats.

We think ten years from now this metaphor will look dated. In the future, when we talk about web pages our children will look at us like we look at our parents and grandparents when they regale us with stories of the golden age of radio. We will access data on the web through a variety of other mechanisms. We see the issue dividing into two separate disputes. First, there is ownership of the browser and the rendering engines that powers it. Second, there is the heart of the metaphor – the web page.

Browser engines Rendering engines, also known as layout or browser engines are the workhorse component of a browser. They take data, in the form of HTML code, and convert it into colorful, information-packed web pages. This is a relatively straightforward process, but one that still allows different browsers to customize on a common engine. One of the reasons that this is



important to mobile is that these engines are able to bridge the gap between the various web standards and the hardware the browser runs on. Rendering engines execute a fair amount of code but all strive to maintain compliance with the various W3C standards. Historically, each browser vendor had their own engine, but over time many of the vendors began using a common engine and sought other areas to differentiate.

This is a complicated arena with many proprietary and open source engines on the market. However, increasingly the browser makers are standardizing on the WebKit engine, with a few notable exceptions.

WebKit is an interesting case study. It began life as the internal engine for Apple’s Safari browser. In 2005, the company decided to create an open-source project with the WebKit name. They opened the code to outside developers and have worked over the years to merge this version into their computers.

Since then, many companies have adopted WebKit as the common engine for their browsers. On the desktop, the leading WebKit-based browsers are Apple’s Safari and Google’s Chrome. More importantly, however, WebKit has gradually emerged as the default choice for mobile browsers. Almost every pre-installed web browser on a smartphone that ships today is based on WebKit. Essentially, the first dispute we mentioned above, has been fought and WebKit has emerged as the clear leader, but not yet the winner.

There are, however, several notable exceptions. Arguably the most innovative browser vendor is Norway’s Opera. While not well known in the US, Opera has an important browser presence in many markets. They use an internally developed engine known as Presto. Opera is important as they are the largest independent browser company on the market. As providing browsers is their only business, they have generally been a step ahead in delivering innovative features such as tabbed browsers. They have also developed a range of browsers optimized for mobile environments.

Another important player is the Openwave mobile browser. Openwave was once the leader in the mobile web, when the mobile web meant low-res WAP browsers. The Openwave browser remains probably the most widely installed browser on the market today, although the limitations of WAP browsing generate only a tiny share of the traffic of smartphone browsers. Many of the thought leaders in the mobile web today are Openwave alumni, but the company went through several difficult years. They sold their browser to Purple Labs which is now part of the Myriad group, a leader in a wide range of mobile software.

Mobile web browsers are by no means exclusively the domain of incumbent desktop browsers. The leading private mobile web browser is Skyfire. They have developed an optimized browser for mobile environments, which in certain situations involves caching content on their own servers for more streamlined delivery. Skyfire originally built their own Gecko rendering engine but have since transitioned to WebKit.

The most prominent of these exceptions is Microsoft. They continue to build their desktop and mobile browsers on the Internet Explorer and their internal Trident engine. This matters for two reasons. First, if Windows Phone becomes successful (a big if) then Trident will remain important. This will depend on the success of Windows Phone, a topic which we explore further below. This is also important because Explorer remains the leading computer browser and Microsoft’s actions hold considerable sway over the broad development of web standards at the W3C. So there was a major sigh of relief when Microsoft announced it would fully support HTML5. This was an important validation of the standard, but then opened the door for the next wave of the browser wars.



Browser wars Mention the subject of the browser wars in Silicon Valley (or probably Wall Street) and old-timers will get a faraway look in their eyes as they relive those glory days. The browser wars were an integral part of the first Internet Bubble.

For the 80% of our readers who are too young to remember that, the story is worth recounting briefly. The remarkable growth and then IPO of Netscape essentially sparked a great wave of interest in the Internet. It brought capital which fueled its early growth before reaching exaggerated levels. Netscape’s founders had created the first browser and then gone onto to start their own company with a commercial browser model. This was widely seen as a threat to Microsoft’s dominance of the desktop. Futurologists of the time talked about a future in which the operating system would be subsumed and all applications would run on the browser and the operating system would become irrelevant. Microsoft took that threat seriously enough to bundle its own Explorer browser with Windows. Over time, this made life difficult for Netscape and eventually pushed Netscape from the market, but at the price of exposing Microsoft to its antitrust woes which persist to this day.

Despite the eclipse of Netscape, we would argue that the vision persisted and has already arrived. The best example of this is Apple. Set aside the iPhone for a moment, their computer business remains very strong today. True, the iPod and iPhone helped this, but more important was the web. A standalone computer relies entirely on the operating system. Anything that is interesting or useful about a non-networked computer has to rely entirely on the OS. Add the Internet, and suddenly there are many more interesting things to do with that computer. Most people use their computers today for three things – work, e-mail and web browsing. Of these three, only the work element really needs a robust operating system, and even this importance is waning. For all those consumers who did not need the best version of Excel and Word, and wanted a simple to use computer, the Mac was a perfect choice. Apple owes its existence to the Internet.

So much for history. We recognize that the 1990s’ browser war ended with the OS remaining dominant, but it opened the door for other entrants. Some would argue that the OS remains important and the 1990s proved this. We do not agree, we think the browser wars were really just the first stage. The next round, the one looming, could very easily go the other direction.

We are already seeing the signs of the direction that the next browser battle will take and think this time around the dispute will likely be a little trickier to follow. Instead of being fought between two competing browsers in full public view, indications are that this fight will take place in the committee and subcommittee meetings of the W3C. In the introduction we mentioned the dispute over web video. One the one hand we have H.264 a standard which has been designed in public set against WebM an ostensibly ‘open’ standard which has been designed by a single company, Google. The company has claimed noble goals for their effort around WebM whose code base they have donated to an open source initiative. There is some credit due their intentions, but we would also see this as their attempt to tilt the standard in their favor. This is just the most prominent of a whole range of issues to be debated.

What is the Web? Stepping back from the minutiae of standards specs, there is a more fundamental debate taking shape over the very nature of the Web. So far this issue has seemed very minor, something that has sparked a little bit of interest on the blogosphere and very little notice in the wider public. That will change. We see this as almost an existential discussion that could change the way in which we interact with data.



Historically, the web developed as a series of links to blocks of text. Over time, pictures, video and eventually applications were added, but the fundamental metaphor of looking at “pages” remained. Today people open web browsers and look at these web pages, but there is no reason that data has to be organized this way. Increasingly, most data is coming in the form of streams of data – RSS feeds, Tweets and Facebook updates. On mobile, the notion of the browser and web page has given way to the “app” or application. If you look under the hood of most iPhone and especially Android apps, they are essentially bookmarks that take users to web pages.

Over time, we think alternative methods for accessing the web will become dominant. The browser will give way to apps, and apps will give way to widgets, embedded data and host other interface mechanisms. We believe that the apps will likely prove transitory for some unknown, more intuitive interface.

In thinking about this, it is important to define a few terms. We tend to use the terms ‘web’ and the ‘Internet’ interchangeably, this is not truly accurate. The World Wide Web is a big collection of web pages; users access them by making HTML calls into the Internet. The Internet in turn is a vast collection of inter-connected networks that store all sorts of data most of which is not structured as a page. The debate of the nature of the Web is sometimes framed as a fight between HTTP and the Web, or of apps versus pages. We frame this as a spectrum of interfaces for accessing data. In some context, say sitting in front of a computer, the interface will remain the browser and the page. In other context, for example, looking up movie times at theaters near you, when an app probably works best. In other situations, embedded widgets with limited processing power, say a time and weather report in your refrigerator.

All of these things will likely rely on the same set of Web standards, but the device in which they reside will play a large role in determining the content they can offer.

In the context of mobile phones, tablets and other portable consumer electronics, this somewhat esoteric debate takes on real meaning. We recently spoke with a developer who contends that in the future companies will build new software with mobile in mind first, and their PC web presence as a second priority. By his reckoning, his latest project is not a ‘site’, it is an app, and this distinction has both technical and emotional connotations.

All of this gives OS makers considerable leverage, and their decisions will shape software developments for their platform. From where we sit, it is still too early to see clearly how this will all shake out. On the hand, the major players – Apple, Google, Microsoft, RIM – talk about HTML5 eventually becoming the dominant platform, but that standard is not ready. In the interim, we anticipate a broad scramble over the definition of HTML5. To draw a historical analogy, this is like the end of World War One. Ostensibly, both sides called an armistice on October 4, 1918, but fighting did end until the agreement was signed on November 11. In the 38 days in between both sides kept up the struggle to lay claim to a final border line. Everyone knew where the end line was, but kept fighting until the end. In the mobile web, everyone knows the endgame will be HTML5, but instead of 38 days this ‘final’ stage of the conflict will last five years, with everyone struggling to eke out the best terms for themselves.

We can make this a bit more concrete. Today most mobile apps are essentially web pages rendered by WebKit. Developers can already write their apps entirely in HTML5, but we think most app-makers would argue that doing so comes with a performance penalty. For instance, we recently heard of one developer who built their app using HTML5 only because they felt this would make it that much easier to port from iOS to Android. A few months later, the developer completed the project and in hindsight now believes this path was a mistake. Using web tools to write the app did save a lot of time in the initial coding of the app, but the



testing process took much longer than expected. The app ran very slowly and the had to go through and troubleshoot each and every bottleneck. Eventually the achieved acceptable performance, HTML5 worked, but the testing time exceeded the amount of time it would have taken to write a native iOS app in Objective C and separately a native Android app in Java. The believe the original performance suffered as it was translated through WebKit. As we noted above, while WebKit is an open standard, Apple continues to be the leading commenter and contributor to the engine. As such, they have a distinct advantage in setting the terms of the engine to their benefit.

Similarly, we now believe Google and Microsoft have both greatly increased their amount of contribution to the overall HTML5 standard to shape other aspects of the eventual shape of the standard. In many ways, we see this as similar to the way in which the cellular standards developed. In cellular, Qualcomm invested heavily in R&D and was very active in the standards bodies. Their contributions greatly accelerated the development of the standard for 3G, but when the dust settled everyone eventually realized that the standard relied heavily on technologies that Qualcomm had patented. We think no one company will be able to claim the lead in the mobile web the way Qualcomm did in 3G, but we do see the big players doing their best to tilt various sub-standards to their favor.

Going native? Setting aside the debate among specific vendors, there is also a healthy debate about the ability of web standards to ever truly replicate the functionality of native applications. Our argument is that web standards will win because they open up the ecosystem to the largest number of developers. We consider this part of the “democratization” of software. Web tools are simpler, that means more people can and do use them. More people using them mean more developers. More developers mean more apps. This is the virtuous cycle of software development we spoke of before. Our thesis in this report is that no other platform out there achieves that. They are either proprietary operating systems tied to Apple or RIM or HP’s hardware. Or they are too fragmented or too small. Web standards fill the practical void of unifying all the platforms and thus eventually capturing the ecosystem.

Despite this, there are many developers who remain vocal that web standards cannot satisfy their needs any time soon. They prefer to build apps to run “natively”, that is on a relatively low-level operating level. Writing applications in C++ is still the preferred method for writing most apps. Converting these apps to Objective-C for iOS or Java for Android is a relatively straightforward problem, something which they have the tools to perform readily easily.

This camp argues that the web standards are still years away from being ready to completely replace all the other platforms. We readily concede this, but feel that HTML5 is close enough to be considered seriously. However, proponents of native systems take the argument further. Using low-level languages will likely always have a performance advantage over ‘scripted’ languages. Web tools will run through the browser and this adds a further layer. This reality is not lost in the web standards community, and we see significant effort in narrowing this gap. So while native apps will have an advantage, we think that with time this gap will narrow sufficiently that at some point in the next few years these tools will be ‘good enough’ to run processing-intensive apps including games.

“Open” versus “Closed” It is always tempting in situations like this to divide the competitive landscape into two opposing camps – usually in the context of us versus them. In mobile, we think there will be a distinction between running native and running in purely web tools. Apple and Microsoft both look likely to favor the native camp, while Google, Opera and others favor the web-only flavor. However, looking at that list should make it clear that this is not a two-way fight. Apple’s view of the open web is very different from Microsoft’s. In particular, we think Microsoft is much more inclined to the idea of native apps, because a purely web-based



world is bad for their core PC-based OS franchise. By contrast, more than any other company, Apple has benefited from a free, creative, largely uncontrolled web. Google for their part seems somewhat ambivalent about the idea of native apps despite controlling what is about to become the most widely deployed native platform on the planet. Google executives have repeatedly said in public that they view the future of the mobile OS as being a purely HTML5 environment. This would seem to be a large problem for developers looking to get greater support for their Android efforts.

One of the chief complaints from software developers about Apple’s iOS ecosystem is the ‘closed’ nature of the platform. Apple controls distribution of software through its App Store. They also determine the pace of advance of the platform itself. New features and new OS capabilities are entirely decided by Apple, and they keep many functions for themselves through so-called ‘private APIs’. The largest complaint is the somewhat opaque nature of the app approval process required to gain distribution of the App Store. All of this fits Apple’s business model of a vertically integrated stack combining hardware and software, and lets them provide the ease of use which so attracts consumers. Set aside the moral arguments of open versus closed, this approach has one drawback. It leaves Apple exposed to rapid changes in demand for OS features. This is a minor problem for now, as Apple has other levers such as marketing and its first-mover status which lets it dictate the pace of change in the industry.

Google, by contrast, has long positioned itself as providing an ‘open’ operating system. Android, they argue, is based on the core open-source Linux kernel. Anyone can download the Android code base and port it to the hardware of their choice. We have written extensively how this shaves the truth quite a bit. True, anyone can download the kernel, but that still leaves extensive work to be done in developing hardware drivers. Moreover, Google restricts access of its several important elements to those who use a more fully-formed version of the code. To gain access to this code, and to apps such as Gmail and the Android Marketplace, hardware makers have to submit to a series of compatibility tests. Again, these benchmarks are publicly available, but beyond that the process remains somewhat murky. Google effectively limits access to this, if by no other means than not having enough resources to respond to all the requests for access. (For more on this topic, please see our past reports including Digits #22 through #25.) In addition, Google controls all submissions to the Android code base, defying the ethos of open-source software. In short, Android is subject to all the same limitations as iOS in adapting to new trends. Google has managed this so far by updating Android at an impressive clip, with five major releases of the code base in just over two years. Of course, this has led to fragmentation for device makers. Google has reacted to this problem by announcing that they will now limit OS updates to roughly two a year, at most.

We have reached the point that we regard any vendor’s claim of ‘openness’ with a healthy degree of skepticism. We see open standards as resembling a series of traffic lights. Open traffic in one direction likely means closing it off in to some other direction. We would caution against assuming that the most ‘open’ vendor wins, as this is such a loaded term. Android is winning because it is free, that is not the same thing as open.

Move toward the cloud

One of the most interesting off-shoots of all this is how neatly it curtails with other developments in technology and networking. If we look at trends in mobile operating systems, and the growing use of web technologies on these platforms, we see a natural next step. Mobile devices have already become browsers with phones attached. This makes them a natural gateway for consumer access of cloud computing services. And beyond, important enterprise cloud tools.



In this view, the phone becomes an important identity tool. One of the chief benefits of using cloud services is their ability to offer access from any place a consumer can log on to the web. Phones tend to be highly personal devices. People share laptops, but no one lets their phone out of their sight for long. Phones also come with connections to carriers’ authentication services. If the carriers could get better at extending usage of these systems, we think it would greatly facilitate the use of phones into new areas. Unfortunately, the operators seem unlikely to do this soon.

This leaves open the potential for the web services gateways to perform this role. Twitter, Facebook, Google and Apple seem to be moving rapidly down this path. Take Twitter as an example. At the recent South by Southwest Interactive show a number of start-ups launched new services or entirely new companies. Many of these provide twists on social networking services. Almost all of these ask to access a user’s Twitter contacts. Our favorite example, is the Lanyrd app. Log on to this, it gives an excellent schedule of SXSW events and then a list of which of your contacts will be attending the show and which events, panels and concerts they have indicated they plan to attend. There are other services which let users opt-in to sharing their current location. Or other apps, like Beluga (recently acquired by Facebook) which let users send out bulk e-mail messages. Combining all of these allowed us to catch up with many of our industry contacts with very little friction. Most people outside of the Valley’s echo chamber assume that Twitter and other social networking tools are consumer toys or diversions. We actually found them to be immense productivity enhancers. Instead of aimlessly wandering the show floor hoping to run into old friends, we were able to add a number of high-quality but unexpected meetings. None of this would have been possible with a laptop or even a BlackBerry’s e-mail service.

Synchronization remains a challenge One obstacle to achieving this vision will be the basic fact that mobile devices do not always have access to mobile networks. Users on planes, on vacation, or on the streets of San Francisco, will have little to no ability to tap into those cloud services. This problem is the heart of HTML5, the ability to store web data locally on a device and execute code locally.

All of these functions, however, lead to another issue – how to keep content consistent across multiple devices. The easiest way to think about this is to consider a document jointly worked on by multiple people. If one user takes a PowerPoint presentation on their laptop to work on during a 15 hour flight to Asia, how will his changes be incorporated into the final document when he regains Internet access, only to find other changes made by users with wired access. A more consumer-friendly and personal example would be music. We have multiple iOS devices and Android devices now. We have stored music on all of these because a little Los Lobos and Hold Steady can really help block out the screaming infant three rows back as we cross the International Date Line. The problem arises in making sure all these devices have our latest music library. Apple solves this problem by synchronizing iPhone and iPad content to a single computer at home. This is immensely ironic (and frustrating), that the company which has done more to end the era of the PC than any other still requires desktop tethering. The press and blogosphere all believe that Apple is set to launch a cloud service, “iTunes.com”, at some point this year. This seems likely to us, but does little to solve synchronization with Android devices.

Ironically (there is a lot of irony in this space), Google has done more to bring consumers into the cloud than almost anyone else. Gmail, Google Maps, Google Calendar are all ‘Cloud’ services. Yet they have done little to bring some of the most sought-after consumer features to Android. In fact, as we point out in the Android section below, they actually block non-Google compliant Android devices from accessing these services. In our view, the most powerful aspect of Android is the single sign-on, that lets users take any Android device and sign on to all those Google services with a single logon and password. So we think it is likely that Google will also soon launch a movie and music service.



We define synchronization as an ‘obstacle’, but the flip side of that definition is that it is a huge opportunity for some other company to fill this void. We see companies like Amazon, Facebook and Netflix all playing a potential role here in synchronizing content. Another company whose product we use regularly is privately-held, Double Twist, often described as “iTunes for Android”, and serves that role admirably. On the enterprise side, privately-held SlideRocket fixes the PowerPoint dilemma we mentioned at the beginning of this section. This is another service we would highly recommend.



OS landscape In the section above we detailed our vision of how the web will develop. We turn now to the state of the mobile web today.

In the Figure 9 we detail the market share of the various OS vendors. This story is now familiar. Symbian has gone from the dominant vendor only two years ago to a much smaller share today. As we write this the press is filled with the latest third party data showing that Android shipments actually overtook Symbian shipments in December 2010.

We have taken a stab at estimating market share going further out. A few caveats on this data. First, our iOS data includes iPhone, our estimate of iPod Touch shipments and iPad. Too often Apple’s share is only counted as iPhone, but the other devices matter greatly to the relevance of the platform. Second, we are likely undercounting Android as this user base is growing rapidly and proving hard for anyone to track accurately. This is especially true for forecast years as Android is getting more widely deployed in a whole host of other devices including set-top boxes, ATMs, and jive-talking Ninja robots. We could easily imagine looking back on this data five years from now and seeing these numbers as laughably low. And should Android get adopted in M2M devices it would make the other numbers a rounding error. And as long as we are dealing in the realm of science fiction, we should note that the other platforms, such as Microsoft, RIM or webOS, could conceivably become smash hits that grow far more rapidly than we expect. In the following sections we detail why we assign a low probability to such a future.

Figure 9: Smartphone OS market share

RIMM15%

PALM/webOS2%

Apple19%

Windows Mobile

10%

Symbian & MeeGo

45%

Android1%

Linux/Other8%

2008

RIMM17% PALM/webOS

1%

Apple27%

Windows Mobile

0%

Symbian & MeeGo

33%

Android17%

Linux/Other5%

2010E




Figure 10: Our view on future OS market share

RIMM15%

PALM/webOS0%

Apple27%

Windows Mobile

8%

Symbian & MeeGo

8%

Android41%

Linux/Other1%

2012E


Figure 11: Our longer term view of smartphone market share

0.0

200.0

400.0

600.0

800.0

1,000.0

1,200.0

Android

Apple

PALM/webOS

RIMM

Windows Mobile

Linux/Other

Symbian & MeeGo


As we have noted throughout, we believe Android will rapidly become the most widely installed smartphone OS, but we do not believe this market share dominance will lead to commensurate industry dominance.



Installed base

So far the focus of much of the analysis on the smartphone space has centered around market share and unit shipments. We think there is a more important metric, installed base. This is a better indicator for developers’ addressable market. In coming years, we expect to hear about new platforms vying for market share, and this certainly matters to the handset and component vendors, but does not address the monetization potential for apps and software.

We actually think this metric explains a lot about the success of the various platforms to date. While we think that with time, web platforms will alter the pattern of hardware dependency, for consumers today there is still a fair amount of lock-in. Users who have bought apps on one platform want to use them when they upgrade their phones. This affect was not lost on Apple, but comes at the price of requiring synchronization with a computer. A big advantage of Android is that its platforms hook into Google’s various cloud services like Gmail and Maps. (We discuss this more in a later section). As a result, we have seen a dramatic shift in user loyalty over the years. One survey we have seen indicates that 86% of current iPhone users plan to purchase an iPhone as their next phone. We were actually surprised that this figure was not higher, but it turns out the survey was conducted in the middle of the “AnntennaGate” press frenzy. We suspect the upgrade rate for iPhone users is probably over 90% loyal. The next closest two platforms were RIM and Android, both close to 50%. This is a huge shift in the industry. Our sense is that in the feature phone market 75% of subscribers switched brands when their two-year contract expired. Nokia probably once enjoyed much higher loyalty rates, but these have faded sharply in recent years, particularly in India and Europe.

Another factor that often gets overlooked is the distinction between iPhone and iOS. We think the iPad and iPod Touch need to be included in this calculation as they are an equally viable part of the developer addressable market.

A note on this data. There is very little hard information available on this subject. At their recent strategy briefing Nokia claimed that the Symbian installed base was only 100 million. This is shockingly low for a platform that has sold about 450 million units over the past six years. For the other platforms we made our best estimate using unit estimates for Apple, Android and RIM from our colleagues, Chris Whitmore’s, Jeetil Patel’s and Brian Modoff’s, company models. We then made some assumptions on what percentage of sales go to existing users who are upgrading to a new phone on the same platform. This is roughly half of the prior year’s installed base as many people upgrade after a two year contract. We turn down the upgrade percentage with time as more pre-paid subscribers join the platform without contracts. We also make some assumptions on the number of users who churn away to other platforms. The churn rate is still relatively low, we used 5% to be conservative for Android, 2% for iOS and 10% for RIM. For Apple, we assume 50% of their reported iPod sales are iPod Touch.

Finally, in anticipation of criticism from the blogosphere, we caution that these are rough estimates which we are happy to upgrade as we get better data. In particular, the size of the iOS installed base is highly controversial. We think this is about 120 million users currently, but we have heard estimates that range from 100 million to 165 million. In addition, we are comfortable with our current Android estimate, but the forecast number could prove highly conservative as growth for that platform is accelerating.



Figure 12: Installed base by smartphone OS

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2009 2010E 2011E 2012E 2013E 2014E 2015E

Contender

Symbian

RIM

iOS

Android


Figure 13: Installed base by smartphone OS 2009 2010E 2011E 2012E 2013E 2014E 2015E

Android 4 52 166 295 436 585 831

iOS 69 121 171 222 272 337 409

RIM 42 63 85 110 129 148 176

Symbian 175 200 170 108 63 25 10

Contender* 0 0 5 27 61 94 137Source: Deutsche Bank *By Contender we refer to platforms such as WP7 or webOS whose precise outlook is difficult to forecast

A few key findings. First, iPhone is still the largest platform. Even if our estimate is too high, the most conservative estimate of 100 million is tied with Symbian. We would argue that Symbian is no longer a viable platform. Second, there have been many press reports about Android unit share overtaking iOS. By our math, the platform will tie iOS for installed base this year. In some scenarios we have run, Android could have 1 billion installed users by 2015.

Also, note that we have included a placeholder for a “Contender”. As noted throughout the report, we think there is still potential for someone to carve out a niche position. This math shows how difficult it will be for that to reach viability. For someone like HP webOS or Nokia/Windows Phone, they are essentially starting from a user base of zero today. We think this year will see very small volumes for either. Barring a design miracle, 2012 will remain an early growth year. By that math, by 2015 any new platform will still be below where iOS is today. Admittedly, Android grew much faster, but that is on a largely unencumbered OS platform that can work on any number of hardware platforms.

It is also worth pointing out that many of these subscriber bases are larger than a carrier’s. For instance, the iOS installed base is already larger than any single US carrier, and by 2015 Android’s installed base will be larger than China Mobile, the world’s largest carrier. The exact impact of this is unclear, but by then we think it will be apparent that the relationship with the consumer will be owned by the platform vendor more than the carrier, as so much of the user experience will be driven by that platform.



App Store update

One of the most widely cited metrics in comparing developer ecosystems among the various OS vendors is the number of apps in each app store. These numbers are all somewhat controversial in that each vendor defines things differently. Apple has tight control over the iTunes App Store and the number of apps can be tracked through that. HP webOS has public APIs which allow for counts to be made. The other platforms are more difficult to track, but each has periodically released app counts. In a recent blog post, Microsoft also made the case that the other platforms over-count apps by counting each language version of an identical app or counting both the “Lite” freemium version and the full paid version of an app. Finally, the Android Marketplace is largely uncurated so many apps there are of low-quality or outright copyright infringers which results in the app count periodically being revised down.

Counting apps is a troubled metric. Many argue that quality should matter more than quantity, but quality is, of course, impossible to quantify. We think the size of the app ecosystem does matter. Each incremental app may add very little value, but the long-tail is important. At the margins, the phone with the Tuvan Throat app wins the one subscriber who wants it. More apps brings a larger universe of potential consumers, to say nothing of the network effect of the first Tuvan Throat singer telling everyone else in his chorus about the great phone he just bought.

We think there is a critical mass needed for each app store to remain viable. One industry contact we spoke with believes the bare minimum is about 2,500 apps. This covers the majority of app usage. So every platform needs its Pandora, Facebook, mapping and YouTube apps. Beyond this is the battle for the long tail. Apple and Android have clearly built platforms that reach this scale. Microsoft has passed the first level, and arguably so has HP’s webOS. RIM’s QNX is seeking to reach both goals by including every possible OS variant including Android on their platform (in the section below we detail why that may not work out as planned). In between these two levels of 2,500 and 100,000, is a big gap, and merely reaching the minimum level is not sufficient as evidenced by Nokia’s Ovi store and the original BlackBerry OS App World. In the Figure 14 we detail the latest available store counts.

Figure 14: Application Store – applications available as of latest date Vendor Store Apps Date of latest data

Palm App Catalog 5,056 1/3/2011

RIM BlackBerry World 25,000 9/10/2010

Android Marketplace 200,000 12/31/2011

Apple App Store 300,000 10/17/2010

WinMo Marketplace 11,500 4/1/2011

Bada Store N/A N/A

Nokia Ovi Store 30,000 1/28/2011Source: Deutsche Bank and company data

As in all things mobile app-related, Apple remains the benchmark to beat. In the graphs of the following pages we compare growth in the number of apps available on the iTunes App Store to growth in the other platforms. We scaled this to compensate for days since launch to make comparisons a bit easier. By this metric, in the first year, Android kept pace with Apple. Microsoft, by the same comparison, is lagging behind, but they use a more conservative accounting and are selling into a much smaller user base.



Figure 15: App Store growth, from launch to one year, number of apps

0

10000

20000

30000

40000

50000

60000

70000

3 60 103 189 249 286 332

iPhone

iPad

WP7

Android


If we look further out, however, Apple appears to be building a bigger lead. In the Figure 16 we extend the data series from launch out to 855 days. In late 2009, the App ecosystem entered the popular imagination and growth accelerated sharply. Android apps fell behind, but have caught up more recently. One cautionary note, the Android data for 2009 is incomplete, and the company had to revise down earlier market estimates when they removed several thousand low-quality apps. We do not think that Android app growth accelerated quite so sharply. We think growth there is accelerating, but is probably not “hockey-sticking” to this degree. As we detail in the section below, the iPad re-ignited growth in the iOS platform, a rate which appears to continue today.

Figure 16: App Store growth, from launch to 2.5 years, number of apps

0

50000

100000

150000

200000

250000

300000

350000

7/1/

2008

9/1/

2008

11/1

/200

8

1/1/

2009

3/1/

2009

5/1/

2009

7/1/

2009

9/1/

2009

11/1

/200

9

1/1/

2010

3/1/

2010

5/1/

2010

7/1/

2010

9/1/

2010

iPhone

Android

Source: Deutsche Bank. Note: Android data between October 2008 and December 2009 is incomplete.



Security

We also think it is important to consider security implications. The industry is starting to awaken to the security risks posed by smartphones. These are threats both to the smartphone users’ data from malware, as well as the threat to the carrier networks in the form of botnets and DDoS attacks. The industry is just starting to look at this, and we imagine it will not be too long before the public and the regulators start to pay more attention as well.

As smartphones grow in importance as a computing platform, they are also growing in prominence among the Black Hat Hacker community and cyber-criminal gangs. Security has been a concern for years on PCs, and the range of malware out there has become very sophisticated. By comparison, the smartphone industry seems to be somewhat of a babe in the woods, taking very preliminary steps in the dark.

One prominent example is RIM, maker of BlackBerry. BlackBerry is highly regarded for offering a very secure mobile e-mail solution, and this is a cornerstone of their sales to corporate IT. However, in recent years RIM has had to move into the consumer space which has exposed it to new risks. Last year they acquired Torch Mobile to bring an HTML browser in-house. Last month, RIM sent out a warning to users that the new browser has a security hole in its JavaScript engine. Details are scarce, but we believe this hole potentially allows access via a remote BlackBerry to the BES e-mail server which sits behind corporate firewalls. If even RIM is struggling with this issue, we think everyone else will have to get up to speed quickly.

Apple is to some degree shielded from security threats. No phone has become as targeted to security threats as the iPhone given its high profile. However, iOS devices have the benefit of tight integration with hardware. This makes it difficult to hack. Further, Apple’s ability to directly update the OS means they can quickly respond to any security threat. We think the security threat on ‘jailbroken’ iPhones has become serious to hold that trend in check to some degree.

Android is a different story. Its “open” nature means that everyone has access to the source code and can scan for security holes. The fragmentation of the OS, particularly along carrier and OEM lines, means that Google cannot quickly supply patches. We believe many corporate IT departments are reluctant to adopt Android for enterprise use for this reason. True, there are third party apps such as Good Technology which can provide encrypted e-mail access, but that security can only protect the app. To date there are no whole-phone security solutions.

One company that is aware of this problem is Motorola Mobility. Earlier this year, they acquired 3LM, a privately-held company founded by ex-Android team members. We have heard positive things about their approach to security, but so far it has not been widely available on handsets in the broader marketplace.

The traditional security companies like Symantec and McAfee have been working on the problem as well. For more on this we would refer you to Todd Raker who covers the security space.



Cheat sheet How to challenge a giant

In this report we argue that Google’s Android has become the industry leader. However, our core thesis is that while Android will soon hold overwhelming numerical superiority, it will not enjoy strategic dominance as Windows once did in the PC. This is a major difference between the PC and the mobile software landscape. We think there will be room for many ‘niche’ players to co-exist in the smartphone landscape for many years.

Despite this, Android’s momentum is very strong. Potentially, it could become large enough to truly warp the landscape to its favor in some more powerful ways. As such, we think many of the competitors in the space are looking for ways to slow that momentum down. No one we have spoken with truly believes they can overtake Android in numbers any time soon, but at the very least many are looking for ways to slow it down. Beyond that, who knows what the future may hold.

As such, we think any discussion of mobile OS landscape often ends up looking for flaws in Android’s deployment. We think the OS has four key vulnerabilities:

1) Fragmentation. Android has numerous fault lines which present headaches for all stakeholders. Developers have to spend significant resources testing their apps across the various Android versions. Handset makers have to re-write drivers and port them to various Android implementations. Carriers worry about being left behind with older phones running outdated versions of the OS. We expect this to get more pronounced as other web companies launch their own variants of Android with their own web services replacing Google, but running the same Android code base.

2) Developer Monetization. Developers regularly lament that they cannot make money on Android. We believe the root cause of this problem is the deeply flawed Android Marketplace. We detail these issues later in the report. The most telling statistic is that over two-thirds of Apps on the Android marketplace are free, while only one-third on other platforms. There seems to be little point in charging for apps on the marketplace as consumers have shown little interest for signing up for the various payment mechanisms. This will incentivize other companies, such as Amazon, to fill that role by launching alternative marketplaces that further fragment the ecosystem and tilt the OS in their favor.

3) Security. This topic has become a recurring theme in the blogosphere recently and could erupt into the mainstream media at any time. The un-curated nature of the Android marketplace has allowed countless low-quality apps to appear. Many of these are altogether fake apps or copycats, but a growing number appear to be malware that seek to corrupt the phone or steal personal data.

4) Litigation. Google is already being sued by several competitors. The most prominent of these is Oracle. Further, Nokia and Microsoft have highlighted that their recent strategic tie-up has a significant patent pool built into it. Should any of these opponents gain leverage through this process, they could slow Android’s momentum by requiring a per-phone license and registration. This process would likely include audit requirements and any such process would likely scare away potential licensee with limited legal resources.



Individual OS review



Android Android unit shipments likely overtook iPhone shipments in December. Most people agree that Android will outship iPhone this year, and we could argue that Android may outship all iOS devices (Phone + Touch + Pad) this year as well.

We are often labeled as Android skeptics, but we do not see ourselves that way. We like much of what Google has done with Android. They have a great OS that rivals Apple’s iOS in many ways and exceeds it in some. But we have also written in detail about all the flaws in the execution of Android - this is likely the source of our reputation. Looking back on the year, Android has had a tremendous run. Consumers everywhere are looking to do more with their phones, and for most of the world, there are still really only two choices for smartphones –Android and iPhone. The OS has considerable momentum now, and that is unlikely to change. Android sales should continue to accelerate, but we have long had a sense that Android disappoints.

We have struggled to define that disappointment. For some, it has been the sense that they betrayed the ethos of open source software. The code base is available to all, but can only be changed by Google, not the community. This has gotten more pronounced in recent weeks as Google has announced they will not release the source code of Honeycomb and moved to make handset vendors sign contracts that limit their ability to customize Android. We see this as an inevitable offshoot of Google’s need to reduce fragmentation and maintain the pace of innovation. This is a fine balance, and Google deserves credit for keeping the innovation coming out at a rapid clip.

For others, there was hope that Google could act as a counterweight to the carriers. Unfortunately, the reality of the business is that carriers still drive phone sales and Google has had to accept constraints on the OS in order to popularize the OS and get it distributed.

In our view, the real trouble is that fragmentation has now become commonplace and as a result, Android is now unlikely to become the dominant OS for mobile phones. Most people assume that mobile operating systems will trend like the PC, with one OS eventually becoming the de facto standard just as Windows did in the PC domain. We think that is unlikely. As we noted a few issues back, developers’ views of Android split depending on the type of app they are developing. For web-based apps, Android is very good. For game-centered apps, like games, Android is still a source of frustration, with fragmentation becoming a major headache.

This is where the comparison with the PC breaks down. Windows runs pretty much the same on all hardware. Android runs differently depending on the phone. In the PC world, developers all flocked to Windows eventually because that was where the users were. This become a virtuous (or vicious depending on your viewpoint) cycle. A larger user base attracted more developers. More developers meant more software, more software drew more users.

In smartphones, things are different. Developers do not see Android as a single platform, as there are many versions of it. This fragmentation is here to stay and as a result developers will likely treat Android differently. As long as they have to port their code to multiple versions of Android, they will view it as multiple platforms. And as long as they see multiple platforms they will think about other options to write to.

Unlike the trend in the PC world, there is no single platform. Android is close, but not close enough to spark the same virtuous cycle.



This should leave room for Apple even as Android swamps it numerically. But we think it could also leave room for other platforms to build their own niche. RIM for one has the potential to build on its large installed base if they can get their new QNX OS ported to handsets and if they can then convert existing users. Not easy, but by no means impossible. The same holds true for Nokia and Microsoft if they can hurry and get Windows Phone ported to lower cost handsets. But as argued in our sister publication, S2N #372 “We are all Smartphones Now”, time is not on their side. There is even room for HP Palm to build something beyond the micro-niche they currently occupy.

We have long argued (Digits #20 “The Big Idea”, March 2010) that eventually smartphones OS will eventually give way to web standards. Phones will become sophisticated browsers and the idea of the browser itself will likely give way to apps as a means for accessing data and services in the cloud.

We are already a long way towards that vision. Most iPhone apps are little more than bookmarks to take you to websites. The key limitation here will be that by definition phones go everywhere, even places not covered by the network. This will mean that a web tools-based OS of the future will still need to have capabilities that access the local processing power of the phone.

Bringing this back to Android, there is a clear split in developer affection towards the platform. In the last Digits, we noted that developers who provide apps for web-based services are all very positive on Android. It is a great development environment for services such as Pandora which does most of its processing work on a server in the “cloud”. Another example is the hit game Angry Birds. Recently, Angry Bird’s reign at the top of the App Store league table was toppled by Bubble Ball. Bubble Ball was developed by a 14-year old. As of this writing, Bubble Ball has generated something like 8.5 million downloads. Of these, about 8 million were on iOS, only a small number of people paid to download this on Android. In a recent note, we pointed out data from analytics company Flurry and Distimo. These showed that about a third of apps available on iOS were free, while on Android about two-thirds are free. Developer after developer has told us that they cannot make money on Android if they charge for their app. As a result, those that make for-pay apps, avoid the platform. This is not a positive indicator for Android.

For developers writing apps that require on-device processing Android has severe limitations. This is true of games in particular, but many others as well. We see four key limitations

1) Android Marketplace. The biggest complaint we hear from developers is that the Android Marketplace remains very weak, and that as a result it is not possible to build a business around paid downloads to Android phones. Several weeks ago Google revamped the Marketplace, and from what we can tell it is generally an improvement but does not really fix the key issues. It is a move in the right direction, and keeps Android competitive, but is not that innovative. A key issue will remain payments, which are still funneled through Google Checkout. So far, consumers have been unwilling to share credit card data with the service in sufficient volume to give the platform momentum. This has left the door open for alternative portals such as Amazon’s Android store and privately-held GetJar.

2) Fragmentation. We detail this in the following section.

3) Performance issues. We continue to hear that performance on Android lags. For example, video frame rates appear a bit slow, audio sampling is also sub-par for high-quality audio features. We think there are several causes to this problem. Not least is Android’s architecture in which code is written for the Dalvik virtual machine, which is a software layer that sits above the core OS. The Java byte code from the VM is then translated by the OS into machine language on a device. Google has sought to address



this through opening up its NDK and adding in just-in-time (JIT) compiler. As far as we have heard, both of these are improvements but neither has solved the bulk of the performance issue.

4) Browser. One of the oddest things about Android is the weakness of the browser. Every Android phone seems to have a different browser, and often they come with two installed. In some ways, this is positive; a key fight on the iPhone was access to enhanced browsers such as Opera and Skyfire. The trouble we find with the standard Android browsers is that they are not that good. Often the built-in Google Android browser underperforms the others. As this note should make apparent, we think browsers will become increasingly important so it is a bit surprising that the Android browser is not built better.

To be fair, all of these appear to be fixable problems with a solution. And we do not want to overstate the issue. Developer interest in the platform remains very strong, driven by the momentum of the overall platform. We point out these negatives because we think they remain very tangible for a large category of programmers. We believe that at heart Google is not that interested in Android as an application platform. Judging from their public statements and their developer relations efforts, we instead see Google using Android as a placeholder until HTML5 and web tools become mature. Google is a web and cloud company, designing device hardware is not their expertise and may not be necessary for their long-term strategic imperatives. However, this stance leaves open room for others to deliver better ecosystems that could potentially draw developer interest and a niche customer base.

Fragmentation

A quick note on Android fragmentation. In past reports, we have commented that the OS fragments along multiple axes. The best known of these is fragmentation by version of Android release. As Figure 17 illustrates, only a little more than half the devices on the market today run Android 2.2 Froyo, with few 2.3 Gingerbread and no 3.0 Honeycomb devices yet showing up. This is a serious issue in its own right as older versions persist far longer than would be expected. There is no official data available for iOS, but there are a few telling proxies. For instance, one developer we follow has posted data on what versions of iOS users of this app carry on their phones. This clearly shows that iOS users quickly and overwhelmingly migrate to new versions of the OS and more importantly, to new versions of devices.

Figure 17: Android installed base by OS version

Source: Google. Used with permission.

The root cause of this lag in Android comes in the form of the second two axes of fragmentation – by device maker and by carrier. One of the great headaches of monitoring this space and to owning an Android phone is the staggered pace of release updates. An



increasingly large share of phones on the market are not capable of receiving upgrades. Changes to the phones’ drivers, either imposed by the carriers or willingly implemented by the device makers, mean that upgrades are a haphazard process. In the blogosphere recently, Motorola caused a stir by announcing that its CLIQ phone will not receive an upgrade to version 2.1 Éclair and remain forever bound in its aging 1.6 shell. This is frustrating, not least because Motorola has promised an upgrade for some time.

Figure 18: Android versions by OS release Platform API Level Distribution

Android 1.5 3 2.70%

Android 1.6 4 3.50%

Android 2.1 7 27.20%

Android 2.2 8 63.90%

Android 2.3 9 0.80%

Android 2.3.3 10 1.70%

Android 3.0 11 0.20%Source: Google

It is unclear how much or if this matters to consumers. Certainly the gadget blogs care a lot, probably too much. Our purely anecdotal sense is that consumers do not care and are for the most part totally unaware of the issue. We think the parties most interested in Google versioning is the wireless carriers themselves. They use Android versions to market the phones (“Get the latest Google 3.0 experience!”). And in many cases the latest apps are not available on older versions of the OS or run poorly. To the extent that consumer notice this, they are most likely to complain to the carrier since there is no one at Google to call.

Conclusion

We see two key areas where Google could cement its success, and two areas where they remain vulnerable.

Android has enjoyed tremendous momentum in developed markets over the past two years. As far as we can tell, the bulk of their volume has come from Europe and the US. This is changing already, and we expect Android’s growth to accelerate as the price of processing power falls. As noted above, these prices are falling to the point that high quality Android experiences are now becoming available to larger and larger swathes of the consumer market. At the moment, low-end Android phones are generally underpowered. When Android was first introduced they highlighted their efforts to make the OS work on some very low power processors, as low as sub 400 MHz processors. As the OS advanced, they have stopped speaking about this. In practice, Android needs a fast processor to be responsive. In this case, time is on Android’s side with processing power rapidly improving. This is one area where the competition is years behind. None of the other merchant operating systems are workable at sub- 1 GHz processors.

Another important advantage is the full range of Google services. In our own experience, we have found the most powerful feature of Android is the ability to do a single sign-on for Gmail, Google Reader, Calendar, Picasa, maps and all the rest. In this regard, Android powers a true Cloud experience. All the Google content we use on the desktop is immediately available on an Android phone.

The irony is that this feature seems somewhat lost on Google. In a bid to control the ecosystem and corral the handset vendors, Google has limited access to these services. Google offers the handset vendors tiered service in the form of API calls to these Google services. Phones that do not conform to Google’s standards do not have access to Gmail, the Marketplace and the rest. And phones built entirely from the freely downloaded open source Android kernel have even more limited access. Users can tap this content directly by using the phone’s browser, but this is an extra step that should not be necessary.



We think this limit reflects a split within Google’s strategy. The Android team itself regularly seems to be headed in two opposite directions. They want to be open, but discourage fragmentation. They want people to use Google services, but limit the access to them. Browsing will be important but they have not invested much time in building one. They believe in web tools but are building an OS around their own somewhat idiosyncratic version of a closed Java-like virtual machine.

This leads to our second concern, the legal front. As with any large company in technology today, Google is fighting a number of standards, regulatory and legal battles. Several of them, like the dispute over web video we discuss above, are going to be important for how we access the web over the long term. Others, however, could have a more immediate impact on Android. In particular, we are concerned about Oracle’s lawsuit alleging that Android is using a number of Java technologies and libraries covered by patents Oracle acquired when it bought Sun and its Java platform. There are ample press reports covering the merits of this case. While we lack the legal background to have an opinion on any of these, it is clear that the legacy of Android and Google’s relationship with Java has overlapped in the past.

Setting aside the merits of the case, we think it is worth a moment to consider the potential impact it would have on the Android ecosystem. The best outcome for Android would be for Google to beat Oracle in court, then there would likely be no changes to the OS. If Oracle wins, or is successful enough to have leverage in a settlement agreement, then we see large potential problems. Oracle could demand some form of lump-sum payment, this would probably be expensive but it is hard to imagine a settlement large enough to be material to Android’s long-term prospects. However, a second possibility is that Oracle requires some form of per-phone licensing fee. Even if this were a nominal amount, we think the impact could be huge. At the moment, Android enjoys its considerable momentum because it is free. Not just $0 but essentially unencumbered. Anyone can go out, download the code base and start building a phone. The requirement to sign up and conform to some form of audit policy that confirms to a standard set by a US court could drastically affect that. And while there are few alternatives, we think the threat of legal action would scare off many of the most enthusiastic users. We have recently heard from handset vendors that Oracle is pursuing a licensing strategy, and is approaching these vendors to sign up at considerable per phone rates.

Summary

In short, we see Android continuing its current momentum and likely growing very rapidly for many years. And while it will likely vastly outnumber other platforms numerically, its flaws and built-in contradictions may prevent it from becoming a truly dominant operating system. Maybe that is a good thing as the mobile web has a better chance of growing the less any single company attempts to control it.



Android – Other than Google These are not the Androids you were looking for

We think this year we will see one more important aspect of Android fragmentation. We think other players will begin to create their own versions of Android. The most likely contender in this space is Facebook, but there are many other potential entrants. The idea here is that third-parties with heavy software expertise will take the freely available Android kernel but alter it to promote their own services and needs. This would save them the time and expense of building an entirely new kernel. Their challenge would then be to ensure that their custom version is compatible with developers’. This is not an easy problem, but we believe it is solvable. And given that Android is already so fragmented it may not make the developer ecosystem that much worse off.

This would not be the first attempt to use the Android kernel. China Mobile already maintains its own fork of the Android code base. They call this OMS and it is the key pillar of their OPhone line of smartphones. To date, this has not been successful and we do not think OMS will ever become a major industry force, but others could learn from its mistakes. The key technical challenge would be ensuring that the code base remains as similar as possible to Main Branch Android. There will always be a lag as new flavors of Android come to market, but Google has said the pace of these releases will slow, so there will be fewer updates to keep pace with. Would-be fork vendors will also have to ensure the API remains compatible with Main Branch Android. Again, Google has already fragmented the OS enough that there is room for flexibility. This is not an easy challenge, but it is also not beyond reason.

The benefits should be clear. In theory, users of alternative Android versions would have access to most if not all Android apps. They would leverage the hundreds of millions of dollars that Google invests in developing Android. They could then focus their energies on maintaining compatibility and porting the OS to new hardware platforms, an area where Google does a poor job anyway.

For over a year, the blogs have been filled with the expectation that Facebook would build its own OS. This now seems unlikely as they appear to be focusing their effort on making sure Facebook works on as many phones as possible including very low-end feature phones. However, we see further benefits to Facebook in going down this path. Facebook’s chief strategic goal is capturing web traffic and the data that generates. Android collects a huge amount of user and usage data and sends this back to the Google Mother Ship regularly. By creating their own variant of Android, Facebook could replace all those links to send the same, equally valuable data back to its own servers. Now to be clear, we have no idea whether Facebook really wants to go this far, we offer them up as an example. That said, there are many other companies who we think will have an interest in doing this and the capability to deliver. This includes other large web companies and even some of the handset vendors who are today dependent on Android for their smartphone line.



Symbian/MeeGo End of an era

The biggest development in the mobile OS landscape in recent months was the news that Nokia will team with Microsoft for its future smartphone platforms. We discuss Microsoft’s prospects in a later section, but felt it fitting that we add a brief eulogy for Nokia’s previous OS Symbian and MeeGo.

The future All we need to say about the future of both platforms is summed up in Figure 19. Symbian will go to zero over some undefined timeline, and MeeGo is “not illustrated”. And probably never will be.

Figure 19: Nokia’s mobile OS forecast

Source: Deutsche Bank and Nokia

The past Since Symbian began unwinding in 2010, a considerable amount of detail about the company’s past has emerged in the press. Former employees may once have been worried about hurting the company, now see that nothing they can say could possibly do any more damage than Nokia has done to Symbian. Freed from this constraint, many of them have published memoirs or spoken to the press in great detail.

Someday, we think the history of Psion and its successor company Symbian will make for a good precautionary business school case study–a study in how to stifle innovation through bureaucracy. One of the key themes to emerge from the turmoil around Symbian is how much innovation the company once had, and in turn how much of it was left to rot by a feuding board of directors.

One of the sad facts of Nokia’s decline in smartphones was that people there clearly knew which way the industry was headed. Years ago they began to focus on software. In our FITT report in 2008, Anarchy in the OS, we detailed ten years of Nokia’s M&A history. While their competitors were busy picking up hardware and chip technologies, Nokia was buying music, video, mapping, and other content and software companies. It appears now that there were two themes to the forces holding back Symbian and Nokia. Firstly, Nokia let its hardware and purchasing teams make all the decisions which hurt software performance. Secondly, Symbian’s Board was split among competing handset vendors, and their conflicting interests put an end to good ideas or hope for improvement.



Nokia regularly touts its supply chain and manufacturing efficiency as one of the reasons it is still important to the handset industry. However, in many ways, those manufacturing assets are an anchor. The company long ago made the sensible decision to build phones around ‘platforms’. These were hardware platforms that were common across a number of handsets, even though the exteriors of those handsets looked different. This is analogous to an auto maker designing a chasis which is then customized for different customer segments. This strategy made Nokia highly efficient in customizing phones to every micro-niche they could find, and the company had anthropologists and design gurus in the field finding new niches. The drawbacks are twofold. Firstly, it prevented them from making major platform changes to customize for large customers such as the US carriers. Secondly, the Nokia platform teams were dominated by hardware engineers and purchasing managers. Often, these decision makers would opt for the cheapest hardware option available. This meant they would scale back processing power or memory to make the phone more affordable. We have heard of many examples where these decisions were made without fully consulting the software teams. The end result was often underpowered smartphones that lacked the ‘snappy’ performance seen in the iPhone and leading Android phones. It also prevented any true standardization of hardware specs. Nokia has tried to persuade software developers that they have the largest installed base of users. But from a developers’ point of view, Symbian was never a single platform, it was hundreds of variations.

Another issue was the history of missed opportunities at Symbian. British IT blog The Register has posted a great history of this. The synopsis of this multi-part history is that Nokia teams were concerned that if Symbian had too many good features it would be adopted by Nokia’s competitors, who also sat on Symbian’s board. They were worried about losing differentiation if it became too viable as a merchant OS. Over the years, they stalled many developments which incorporated the best pieces into their own S60 user interface. The Symbian UI never really developed and it was left to each of the handset vendors to build their own UI. This effectively hamstrung Symbian and negated the benefits of having a merchant OS able to amortize R&D costs over a number of customers. Some good posts on the subject can be found on www.theregister.co.uk

The end result of this was that the phones launched by Nokia had neither the internal organization nor the external resources to respond quickly. Sometimes having the first-mover advantage comes with path dependencies that someday become first-mover disadvantages.



RIM RIM is fragmenting itself

RIM is not yet shipping a true smartphone platform in our view. We count RIM as one of the smartphone platforms, mostly because consumers view their BlackBerrys as a smartphone. In our view the current BlackBerry OS is not really a true operating system; it does not allow much in the way of third-party applications to run on it. It is also underpowered relative to true smartphones. For the time being, this is an academic distinction, one that is lost on all the consumers who are just happy that their phones do e-mail. With time, we think consumers will gradually become aware of the yawning gap between what their BlackBerry can do and what other smartphones are capable of.

RIM is fully aware of this problem. And to their credit, they have people who are at the leading edge in thinking about mobile operating systems. In fact, the first person to ever mention HTML5 to us was an RIM employee. The company clearly understands the direction of the industry and is moving ahead quickly to offer a viable platform.

To this end, the company began building a new OS several years ago. We believe they did much of the initial work with a company called QNX (pronounced Q nix), and eventually acquired the company to bring its design experience in-house. Similar to the Nokia of old, RIM has also been acquiring other software technologies. Two years ago they acquired Torch Mobile to help them build an HTML5-capable WebKit browser. Last year they acquired The Amazing Tribe (TAT), a very innovative design house that specializes in mobile User Interfaces (UI).

At face value, QNX has considerable appeal. The Posix-based QNX kernel has a reputation for being robust and stable, with considerable deployments in industrial and aerospace applications. We also like what we have seen of the architecture. It appears to be somewhat modular, meaning it has the flexibility to support a number of different operating environments. More specifically, we see three highlights. First, it is secure which is important for the core enterprise-centric RIM customer. We think QNX allows for elements of the OS and the device to be sandboxed or securely partitioned so that apps running on the device cannot access enterprise networks. Second, it is flexible enough to support several development paths. True to RIM’s core long-term vision, QNX lets web designers build apps using standard web tools. In theory, this ‘future-proofs’ the OS for an HTML-centric world. A third, somewhat related point, is that the OS appears to be extensible, meaning other development platforms can be slotted in. In particular, this allows for developers to write Flash-based applications and run them on the Adobe AIR platform, which is running on top of QNX. Flash is an important tool for web developers today, and important tool for web advertising platforms. From what we can tell, QNX devices can support Flash videos and Flash apps out of the box. In theory, there is room for other OS’s to run ‘inside’ QNX. For instance, RIM has announced they plan to enable QNX to run Android apps as well.

As always, their trouble will be executing to this vision. We see a few drawbacks here. First, QNX appears to be a relatively ‘heavy’ OS, meaning it requires significant processing power and memory. RIM’s co-CEO was quoted in the press as saying QNX needs to run on a dual-core processor. This means it will not run on handsets today. The first QNX devices will be tablets coming out later this year. RIM needs to be able to port QNX down to handset-sized devices. RIM management has said they will not have QNX-based handsets until 2012, and even then it will likely only be available on more expensive devices. This is a problem for RIM as much of their recent growth has come from emerging markets such as China, Indonesia and India. Within a year, $100-$200 Android handsets will be available in those markets, and RIM will have no ready answer.



A second concern remains the development ecosystem. We mentioned above that the ability to support Flash, Adobe AIR and Android apps demonstrates how powerful QNX is. The drawback is that for developers this is somewhat confusing. Adobe Flash is known for being very processor-intensive. It is unclear what this means for battery life, but we suspect it will be an issue, especially for smaller devices. More importantly, from a developer’s point of view, there seems to be little incentive to write apps for QNX. Developers are still in early stages of writing HTML5 apps, reducing the appeal of that side of the platform. Secondly, while Flash holds considerable appeal for web designers it holds considerably less appeal for application developers. Here we are largely thinking of game designers. While many of these dipped their toe in writing Flash apps, our sense is that there is almost no enthusiasm for designing applications for Adobe Air. The hard-core game development community would far prefer to write apps in low-level languages such as C++ or Objective-C. To address this, RIM has just announced their own Native Development Kit that lets programmers write to Posix kernel. We read this as an indication that their Adobe Air strategy has largely failed, a view confirmed by a developer we spoke with who had warned RIM that his company had no interest in AIR.

RIM appears to be aware of this issue and more broadly aware of the difficulty in building a developer ecosystem. Hence their decision to open up the platform to Android apps. The appeal would be that Android already has a much larger developer base and will likely always have a lead. RIM management has said that by letting Android apps on to the platform, it will address the long-tail appeal of having hundreds of thousands apps available, or as they term it ample “tonnage”. However, the details of execution matter tremendously here. There are two ways to run Android apps in QNX. The first is to run the Android app on a Java virtual machine (VM), or more precisely a Dalvik VM. In theory, this will work because Android itself runs all code on Dalvik and then translates that Bytecode into executable files. In practice, this will likely translate into users running one app to open up the Android environment, then run the specific app on that VM. Essentially an app running on an app. Apple has actually banned this, in part because they want to control the platform, but also because there is a very serious performance penalty involved. A second method would be to run the whole Android OS as a virtualized OS within the RIM device. This requires a significant engineering effort, essentially porting Android to BlackBerry. This will of course take some time. It will also further fragment Android. More importantly, we think RIM would essentially make itself just one more Android handset vendor whose only differentiation is better integration with the BlackBerry e-mail ecosystem. That seems like a strategic dead-end, cutting into RIM’s hardware margins.

If all this seems confusing, it is. The future RIM OS strategy is actually five platforms:

BlackBerry Java Virtual Machine to support the 25,000 existing apps that run on the legacy BlackBerry OS

Dalvik Virtual Machine to run Android apps

Web Works – the web tools-based run-time for QNX

The QNX Native Development Kit (NDK)

Adobe Air

Rim has essentially fragmented itself. This is the same strategy on which Nokia and Motorola foundered over the years. They have to incur large operating expenses to support all of these platforms. They also confuse developers to the point of distraction. From developers’ point of view, why would they learn Web Works on the intricacies of the QNX NDK when they can just focus on Android. This will lead to QNX withering on the vine. In the end, Google will drive the pace of Android innovation and RIM will have to be constantly running to keep up to date with the latest version of Android. On the recent earnings call, RIM management touched on this topic and tried to play down the importance of Android Honeycomb insisting



that developers only care about Android Gingerbread. We think that is backward looking. Moreover, it runs that risk that the next Android update is meaningfully better for developers and it takes months for RIM to provide an update.

Finally, we are concerned that moving to the software-intensive smartphone world may prove beyond RIM’s core competency. While they have many smart people, their history in software products is poor. Beyond the core RIM e-mail experience and BlackBerry Messenger (BBM), we can think of no software product that RIM has executed successfully. BlackBerry Everywhere, Unite, the list goes on. Admittedly, we do not know what the production build of QNX will look like. What we have seen of Playbook demos is not encouraging. The Playbook UI appears to borrow heavily from HP’s webOS UI, with many very similar usage metaphors. More alarmingly, the Playbook’s e-mail application looks almost identical to that on the iPad. We would have thought RIM could come up with a clever interface on what is their core product. Seen in this light, their acquisition of TAT looks to be driven more by necessity than a desire to innovate. All of these leave us questioning RIM’s ability to be thought leaders in the smartphone market and to find ways to differentiate their products.



HP webOS Surprise underdog or just an underachiever?

We continue to hold the view that webOS remains the best mobile operating system on the market. Unfortunately, the execution behind it lags significantly.

From a technical standpoint, webOS has a lot going for it. We think they have developed the most intuitive UI for mobile-sized devices. Under the hood, their extensive use of web tools is considerable, especially given our view that ultimately the whole industry will someday move down this same path. webOS has some of the most robust multi-tasking on the market. Others claim to have this capability but it is either hidden from the user by design (i.e. Apple and Microsoft) or by a confusing UI (i.e., Android). They have also developed some very innovative browser-based development tools that, while unpolished, point to a whole new direction for software programming, not just for mobile but for the broader software universe. The OS has one of the most open distribution models on the market combining a curated App Catalog with a wide open web-based model that essentially lets anyone distribute apps.

One of the chief technical complaints we hear about webOS, however, is that for more complex apps (i.e. games) there are still limitations around performance. HP (and Palm before them) addressed these shortcomings by offering a Native Development Kit (NDK) which lets developers use low-level languages like C++ to write apps that call directly on device hardware. We still hear that the NDK is lacking and is not as fine-tuned as other parts of the OS. Even this, we see in a somewhat positive light, as the company is several years ahead of competitors in dealing with issues that will inevitably crop up as we bridge from native to web-based development.

All that being said, the future of webOS remains darkly clouded, both from technical and market standpoints. On the technical front, webOS looks to be a large but unfinished product. For instance, those interesting design tools we mentioned above appear to be a great idea, but are not yet sophisticated enough to handle full-fledged development and remain a sore spot for developers. Many aspects of the OS look unfinished or not fully fleshed out. Given Palm’s history, the webOS team did a tremendous job of putting out a good initial version of the OS, but clearly lacked the resources to complete the vision. Since HP’s acquisition the OS, they have continued to improve webOS, but still have some way to go.

In part, this is largely due to the provisional nature of webOS’s standing in the marketplace. More directly, the OS has not gone far because they have not been selling that many devices. For developers, we have found very little interest in developing for the OS because there is such a small installed base and thus very little market to tap into. The webOS App Catalog has only about 5,600 apps in it. It has hovered at this level since before the HP acquisition. We use RSS feeds of its App catalog to track develop activity. As shown in Figure 20, this faded sharply as Palm declined, remained low for most of HP’s ownership. Only recently has activity seemed to revive. HP recently announced several new webOS products including a tablet, and this appears to have sparked a bit of new activity, but the base remains very low.



Figure 20: webOS developer activity as measured by RSS feeds of the App Catalog

0

100

200

300

400

500

600

700

800

900

13-Feb-10 13-May-10 13-Aug-10 13-Nov-10 13-Feb-11

Source: Deutsche Bank and HP

All in all, we think webOS remains a viable contender in the mobile OS space, but has a low probability of gaining traction.

Since HP’s acquisition, we think the OS has received an infusion of R&D investment, and from what we have seen they have made some real improvements in webOS. Our sense is that HP has a sincere interest in making webOS work. If successful, it gives them the potential to improve the margin structure of their mobile business by offering a truly differentiated user experience. Moreover, they have talked openly about porting webOS to work on printers, PCs and other hardware. This means they could theoretically have a differentiated offering in the PC space as well. This would be a meaningful improvement over standard margins versus the merchant OS space. For more on this potential we would refer you to our colleague Chris Whitmore’s body of work on HP.

In the mobile space, we think webOS has been challenged by resource constraints. Palm did not have enough to invest in completing all features of the OS. Now they do. More importantly, Palm lacked the resources to market the device and withstand the losses required to build an ecosystem as the user base slowly grows. It is still unclear if HP has the marketing budget and expertise to do this. More importantly, the mobile business will still require carriers to support this product. This is what made Palm uncompetitive. In the end, they could not market the device enough to drive carrier interest. There is no quick solution to this. It will take several years and sustained will for HP to succeed here. So while we are encouraged by their expansion of the product portfolio, we nonetheless feel webOS’s future will remain an open question for several years.



Microsoft Windows Phone The best defense is a good offense

It is almost impossible to have an objective conversation with anyone in the industry about Microsoft. The company has so large a presence in everyone’s lives and such a history in the industry that rational thought is usually impaired by some perception. That is no less true in mobile. A relatively senior member of the Windows Mobile team once told us they wished they could have named their OS “Zune Phone” as a way to get away from the emotional baggage the Windows brand carries.

All of the above is our way of saying that the challenges facing Windows Phone are as much marketing related as they are technical.

The technical side is relatively straightforward. Microsoft Windows Phone 7 (WP7) is a relatively good offering. Like webOS, Microsoft came to the market with a truly original design and user interface, not simply a new tweak of the iPhone layout. We believe the Windows Phone UI has a lot to offer, but is not as immediately intuitive as iPhone. Users as young as 18 months old can look at an iPhone and know instantly what to do. For WP7, the learning curve is a bit longer, it takes anywhere from 30 seconds to a few minutes to fully grasp the interface. However, after that point we think it is both easy to use and appealing. The company also makes a big deal about the extent to which it is much more customizable than iOS; a facet which is interesting but probably gets overplayed.

Under the hood, the OS is relatively robust and offers a comprehensive set of APIs. The SDK exposes most of the functions developers need, including low-level access to device hardware. The catch is that developers really need to use Microsoft tools to program it. This includes experience with Silverlight and other assorted elements of the Microsoft graphics stack. This would be somewhat limiting for a brand-new OS, but given the fact that Microsoft’s developer base is already huge, there is sufficient pool of engineers comfortable with these elements.

This is particularly true for games. Microsoft has been working hard to encourage developers to write apps for WP7. “Working hard” is a bit of a euphemism. They have been aggressively courting small developers of popular apps, offering significant upfront payment for projects. On the gaming side, we think they have been “encouraging” existing X-Box developers to write games. The end result of this is 11,500 apps on the Windows Marketplace for Mobile, a good showing for an OS that has been on the market for less than a year. Perhaps most impressive is the range of quality games available. We say this not only personally as game aficionados, but in comparison to the poor showing of gaming titles on Android. We think gaming is a very important app category as it is one of the easiest ways to introduce users to app downloads.

So far so good. In characterizing our criticism of WP7, we keep in mind that it is still a first-generation release; however a lot of work still needs to be done for developers. This includes a still cumbersome app submission process, limited developer support and a range of uncompleted minor features. Apple went through this process, Android is still struggling with it, and those OS are now into their fourth and third release, respectively. The market has been waiting for an update to WP7, and the blogs were filled with expectation of when it would arrive. The first update is set to come out in the next few weeks. We would have expected some updates by now, and are a little surprised that it took this long. A month ago, Microsoft pushed out some form of software patch, which seemed to be a stepping stone to a full update. This had the dual negative effect of being less than expected and causing a



number of Samsung WP7 phones to be transformed into large black paperweights. Since then we have now seen a true update, including such sought after features such as “Cut & Paste”. Microsoft had said that many of these features should have been ready a month or two after launch. Since we have not seen them yet, we have to wonder if the announcement with Nokia has distracted Microsoft’s efforts. Instead of focusing on updating the OS, they are now focused on getting the first Nokia device to market. More on this below.

Do consumers care? The chief challenge Microsoft faces is the lack of consumer interest. The company said they have shipped 2 million WP7 phones, but have so far refused to comment on activations. Our sense is that there are still a lot of WP7 phones sitting in carrier warehouses. Despite spending a huge amount of money marketing the devices (see Digits #23 August 2010), consumer awareness remains negligible. Microsoft management has repeatedly claimed their biggest challenge is just getting consumers to try the phone, because once they try it, they will like it. We agree with this assessment, but caution that they still have a long way to go.

Notably, Microsoft’s recent television ad campaign received a rather lackluster consumer response, and we believe it has since been scrapped. The chief problem is defining what makes this phone so different that consumers should care. There is no easy message to send. We believe consumers have a hard time distinguishing smartphone operating systems, and defining what makes WP7 is not clear. “We offer tiles” has little resonance despite being a nice feature, while “we are more personalized” or “we have better Facebook integration” are tough to prove.

So while Apple was able to leap to success through diligent application of their marketing magic, Microsoft will likely have to slog it out, slowly building up their user base. We would caution that any other company lacks the staying power to do this, but our sense is that Microsoft has both the persistence and the resources to stick with their efforts over several years. The stakes are too high for them to abandon this space, and they have the finances to keep at it.

Prospects The big recent news for WP7 is their February announcement that Nokia plans to build its future smartphone line around this OS. This is a big win and a prominent topic of conversation at the recent Mobile World Congress. However, without this announcement the outlook for WP7 would be considerably darker. For better or worse, the future of the OS is likely dependent on Nokia. The trouble with this is that we think it could be some time before Nokia actually brings out any products. As we noted in the March 1 Digits from Barcelona, we think Nokia will have its first WP7 product on the market late this year, but this will be built by an ODM partner (likely Compal). The first, truly Nokia-built devices will not be on the market until the same time next year. That means that fully five years after the launch of the iPhone, Nokia will have its first response. Not encouraging.

Another challenge is that Microsoft is at heart a software company that works to software timetables. In this world, release dates are fungible, and if anything does not work it can be fixed in a follow-up patch. Microsoft can and has pushed out release dates. In mobile, a late ship date results in missing hardware product cycles. OEMs need to have code in time to physically build the hardware. If Microsoft pushes that software release from summer to fall, the hardware vendors end up missing Q4 product cycles and the key Holiday shopping season. We spoke with one hardware vendor who believes his whole company would have failed had they gone with Windows Mobile 6.5. This gives us reason to be highly cautious of the WP7 outlook for 2012. Even if Nokia is able to ship five models next year, we think the most likely outcome is that it makes only a small dent in the installed base.



This is especially true as it now looks likely that Microsoft has lost most of its other hardware partners. As we noted in our post-Barcelona report, HTC, LG and Samsung are not happy with the preferential terms Microsoft provided Nokia. Crucially, this includes some period of exclusivity for new releases of the OS. This will give Nokia a big advantage in winning carrier business, to the detriment of the other three. Our sense is that Samsung is now actively exploring other options. They may continue to carry a token WP7 product line, but seem unlikely to emphasize it. LG seems ambivalent and will likely behave as Samsung does, and they are now too small to be a major force anyway. HTC has patiently coped with Microsoft’s idiosyncrasies for years, and will likely continue to remain a WP7 vendor.

Another concern remains the cost of WP7 devices. Not only do these phones have to pay for the Microsoft license (which we think is still between $10 and $20 per phone), but Microsoft is very tight on the hardware specifications for its devices. They require that phones use powerful processors, have large screens and considerable amounts of memory. This ensures that the phones perform well, but also means they are not cheap. We think the age of the $100 smartphone is only one to two years away. The average WP7 phone is today $400 - $500 and we have seen no sign that they will close this gap quickly. Hardware partners have told us that they are actively encouraging Microsoft to work down this price curve. And Microsoft itself is aware of the problem. Even here, we think it will take several years before this spirit can translate into action.

Adding to the list of deficiencies, Microsoft remains far behind on tablets. Most people agree that Big Windows 7 is not suited for tablets. The company has indicated that they are holding off tablet support until they can build in all the requisite features for the next major release, Window 8, which is not expected to be on the market until late 2012. Moreover, they seem to be positioning Big Windows as the tablet OS, not Windows Phone. This is a major departure from the rest of the tablet community which views tablets as needing far fewer features than Big Windows offers. Our sense is that when the company planned its tablet strategy it felt that WP7 was not secure enough in the market to fill what could be a crucial strategic role for the company. The price they pay for this view is a multi-year delay in the availability of Windows tablets.

Conclusion Our goal here is to sound cautious but not overly negative. Despite all the challenges we outline above, the fact remains that there are only two merchant OS on the market today – Android and WP7. Every other entrant is tied to a single hardware platform. Microsoft’s prospects could improve considerably if Google stumbles. In several sections of this note we highlight areas where Android is vulnerable. Chief among these are legal challenges.

Microsoft has been very vocal on this front. They have consistently said that “Android is not free”. At inception, we interpreted this message to mean that Google does not provide much support or device drivers to the hardware community building Android phones. This support work is expensive and not always apparent.

Recently, however, we think Microsoft’s “Android is not free” message has taken on a new tone. An important pillar of the Nokia-Microsoft deal is the pooling of their patents. Both companies are already suing many members of the Android community. Microsoft has made it clear that they plan to “protect their IP”, and we clearly interpret that as meaning the best defense is a good offense. If Android phones become entangled in complex litigation or worse begin requiring payment of a licensing fee, we think that could greatly dampen smartphone users’ enthusiasm for it. Several companies are closely examining the Android code base. These companies are offering hardware community comprehensive licensing coverage for Android phones, but often are per-phone payments in excess of roughly $25 that Microsoft is asking for a WP7 license. We have no ability to gauge the merits of these legal issues, but caution that should any of them succeed it could dent Android’s momentum and open the door for Microsoft to win new interest.



Update From our examination of the available data (e.g. from windowsteamblog.com), we note that apps are being counted differently on different platforms. For instance, on both Android and iTunes, different language versions of each are app are counted individually. Also, many apps platforms have free “Lite” or “Trial” apps which are freemium versions of apps that encourage users to pay for full versions. WP7 has a Trial feature built into the OS which eliminates the need for this kind of versioning. Based on the available information we think it would be fair to say there is a danger that other platforms ‘overcount’ the multiple versions of many apps, and this can greatly exaggerate the individual number of apps. So while the WP7 Marketplace definitely has fewer apps (and as we note elsewhere is tracking behind the other platforms in app submissions), we think the gap is nowhere near as wide as the raw numbers seem to indicate. There is no easy way to quantify this or adjust for an apples-to-apples comparison, however our sense is that the WP7 Marketplace has reached viability, and is doing better than would be expected given the limited size of the installed base.



Apple It is good to be in the lead

Through all the discussion above, it should be clear that we think the mobile landscape will remain fragmented for many years. Any company could potentially carve out a niche, and Apple looks set to retain the most profitable of those ‘niches’.

Our view is that Apple is unlikely to ever become the dominant handset vendor. They will continue to capture a very profitable segment of the market, essentially taking the cream of the demographic crop. We believe there are several impediments to them moving beyond a 20%-30% share of smartphones. Most importantly is the value of the Apple brand which gets diluted if everyone has an iPhone and starts to feel less magical. This is a controversial view, and many people expect Apple to launch a lower priced iPhone at some point in the future. This could greatly expand their footprint. That discussion is beyond our scope, and we recommend you reach out to our colleague Chris Whitmore who covers Apple. For the purpose of this report, however, we have based our forward assumptions on that upper bound to Apple which still leaves ample room for unit growth.

Developer ecosystem – Apple remains the favorite For a long time, the blogosphere was crowded with complaints about Apple’s somewhat Byzantine approval process for submitting software to the App Store. Ask developers about it today, they will still complain about the process, but that has not stopped them for writing for this platform. The truth is that iOS and the App Store remain the best ecosystem for developers. This combines both a large installed base of users – including iPod Touch and iPad – as well as a trusted payment system. Android will someday have more users, but barring a dramatic change to Android Marketplace the Apple App Store will remain the best place for developers to make money and reach users.

To back up our qualitative assumptions we took a look at the number of apps on the platform. Despite the numerical success of Android, there appears to be no let-up in the pace of growth.

Figure 21: Growth of all iOS apps

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

7/11/2008 7/11/2009 7/11/2010

Appstore

iPad

App-Store ex iPad

Source: Deutsche Bank and Apple



The graph in Figure 21 shows the number of apps continues to grow at a very healthy rate. There was a slight pause in 2010 as developers shifted toward writing iPad apps, but then overall growth took off again after a few months. If anything, the launch of the iPad has accelerated growth of the ecosystem. In the graph below we compare growth of iPad apps from launch to today with the number of apps in the App Store in the year after the initial launch of the App Store for iPhone and Touch. iPad app growth has outpaced the original growth curve by a considerable amount. Admittedly, the iPad enjoys the benefit of a mature platform while the original App Store started from essentially nothing. Nonetheless, the pace of growth for the App Store should be encouraging news to Apple.

Figure 22: Number of iPad vs. iPhone apps, days after launch of App Store

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

3 60 103 189 249 286 332

iPad iPhone

Source: Deutsche Bank and Apple

Conclusion In all of our research for this report, including extensive travel to speak with people throughout the Apple ecosystem, it was tough to find a single negative data point about the company. The skeptic inside us knows that glory fades, and no one has ever stayed on top of mobile product cycles for more than a few years. Despite this, we can find no sign of any decay in Apple.

Their chief challenge is to keep the pace of innovation and awe going. Apple’s success rests to a considerable degree on their design abilities and their marketing prowess. This is difficult to master indefinitely. Setting aside any non-market factors, like another macro shock or a change in CEO at the company, we think Apple can keep this pace going for several years to come. We think they are continuing to advance the platform. This will likely include updates to the iPad and iPhone this year. Apple’s cadence to date has been to change the iPhone form factor every two to three years. Our guess is that the iPhone 5 will come out later this year with modest changes to the iPhone 4, followed by an iPhone 6 next year with a radically different form factor. This should be sufficient to keep users loyal to the platform through 2014 at the very least. In addition, they could potentially further expand the franchise in several ways. As mentioned above, they could roll out a new low-cost form factor, although we think this is unlikely. They are also gaining considerable traction in the enterprise as large corporations are increasingly opening up e-mail to devices other than BlackBerry. Furthermore, the industry expects Apple to roll out cloud services for media at some point this year. This would further enhance brand loyalty and cement the platform’s position.



Adobe Flash We close this section with a look at Adobe’s Flash ecosystem. We will not spend much time on the topic as we feel its future as a substitute for a mobile OS is limited. Nonetheless, we have fielded a large number of investor questions on this topic.

The Adobe story was once very straightforward. They were the company that made digital content viewable. This began with text documents in the now ubiquitous PDF format. Then, they spread their wings to web page tools. Finally, through the acquisition of Macromedia, they made video content viewable on the web. The Flash protocol has become the de-facto standard for viewing video on the Internet. Flash filled a hole in the original Internet. At the time, there was no standard for viewing video on a website, and HTML offered a practical solution to the problem.

Through this, Adobe built a very strong following among web designers (not developers) who learned to code Flash for many web pages. Adobe sought to build on this success by expanding the functionality of Flash. On the desktop Internet, Flash became a way to create all kinds of graphical content from fancy effects on web pages to executable applications such as games.

The appeal of Flash on the desktop Internet is that it lets web designers write code that works on any hardware so long as the browser supports Flash. Since video was so important, all browsers eventually built in this capability. Given our thesis, that eventually web tools will come to dominate the mobile landscape, it would seem natural that Flash has a role here. Unfortunately for Adobe, this does not seem to be the case.

The first hole in the armor came with the lack of Flash support in the iPhone. This was followed by the very public dispute between Apple and Adobe on the topic. Apple claimed that Flash was poorly executed, a processor hog and a major drain on battery life. Adobe countered that Apple was just seeking to protect its control of iOS and did not want to be threatened by alternative execution environments. If Apple had been the only hold-out, then it is possible that Flash could have remained an important part of the mobile web and Apple would have been forced to eventually build it into iOS.

As it turns out, Flash appears to have several other flaws. Firstly, there are clearly performance issues associated with Flash. While we do not think Apple’s claims are fully accurate, we have seen signs that Flash carries a power penalty, a major issue for smartphones. Secondly, Adobe has struggled to make Flash available on all other smartphone platforms. Getting Flash to work is a time-consuming task that requires very low-level integration work. This appears to be one of those problems that just takes time and is not solved simply by throwing more people at the problem. As a result, Flash availability on every other OS is a one-by-one process. For instance, some Android phones support it, most do not. A third problem is that Flash appears to not live up to its claim of write once, run anywhere. Since this is the cornerstone of its appeal, this is a major problem. Developers have told us that Flash programs need to be customized not only on a per-device basis, but on a per-chip basis. This means apps will need to code to accommodate each version of a Qualcomm baseband, for instance. This is not only beyond the scope of most web designers, but would seem to cripple any hope of future-proofing apps. With time, we think Adobe could sort out these issues, but the web community seems to be moving on without them. Our sense is that HTML 5 support of video is making significant progress. While we note several areas of debate in sections above, our sense is HTML5 video has sufficient critical mass to remove the need for Flash in mobile. And any debate on that subject looks likely to be resolved before Adobe can make Flash truly viable.



Beneficiaries One of the great frustrations in researching this space is the fact that there are few direct investments to make based on our findings. Apple certainly looks set to do well, but few other stocks offer as direct a correlation. It is tough to justify buying Google because of Android. We estimate each Android phone generates about $10 per year in search revenue for the company. That is a good number but a drop in the bucket for the search giant. Even with our relatively optimistic forecast to how large it could get, it looks likely to be many years before this is a $1 billion business. Buying Microsoft for Windows Phone exposure makes as little sense as buying HP for webOS exposure. Buying the handset vendors also looks less promising as many of these companies are likely to see margins compress as smartphone competition heats up. As such, the team is negative on RIM for factors directly related to the growth of other platforms. Surely there must be other ways to participate in the growth.

We think we can identify a few buckets of stocks which do benefit directly these include:

Mobile software vendors. The stock market is littered with less successful companies in the field of mobile software. Few are worthy of attention, but we think browser maker Opera is often-overlooked participant in the mobile landscape that merits more investor attention.

Processor libraries. One of the more popular smartphone investment themes has been to buy Arm Holdings. More powerful phones need more powerful processors and Arm has clearly emerged as the champion of that platform. In a similar theme, NASDAQ-listed, Israel-based Ceva should benefit as well as their processor libraries are showing up in more DSPs and media processors.

In a similar vein, we think componentry for phones has to increase in content as phones become more powerful. This spans the range from audio-visual components to better radios and front-end modules to capture the proliferation of radio frequency bands.

Service providers. Aside from the wireless carriers, whose prospects are beyond the scope of this report, there are many companies which can provide services that work particularly well on mobile. Many of these companies, Internet based, some still private.

Equipment vendors. This topic is also beyond our scope here, and our team covers it well in other reports, suffice it to say there is a pressing need to upgrade network infrastructure at almost every level.

We have already covered the topic of browsers in an earlier section in the note. Below we discuss some of the companies closest to our coverage universe in the component space.

Faster, higher, denser

One of the tenets of our investment thesis holds that hardware is a commodity, and the key to smartphone economics rests in software. This is sometimes misinterpreted as meaning there is no future in hardware. To be clear, we are referring to the hardware and design of phones itself. No handset vendor can hope to differentiate solely on the basis of their hardware design as this can be readily replicated. For the companies that supply that hardware, however, there is an immense opportunity on the horizon. Smartphone vendors are engaged in a race to provide better specs on their hardware precisely because it is such a commodity. If one company has a smartphone with an 8 megapixel camera, others will have to follow suit to prevent that first vendor from differentiating. That is bad for the handset maker, but good for the supplier of 8 megapixel camera modules.



We see two axes of growth. First, the ‘arms race’ in hardware will help leading companies that have a technology or process advantage take a lead over smaller competitors with lesser resources. Secondly, the declining price of smartphones means that more and more phones will need advanced features. More volume and more content per phone is a good combination in a smartphone market that is growing 30%+ a year.

Most of this should be familiar ground to readers, but we have been able to quantify. Over the past two months, we have scraped data from a variety of web sites and other sources with data on smartphones to compile a history of phone specs going back to 1995.

No one would be surprised that phones have gotten smaller over the past 15 years. The average phone volume has decreased from 225 cubic cm to 78 cm^3 today. More interesting, in that same period, phones have gotten denser, going from 0.94 g/cm^3 in 1998 to 1.24 g/cm3 today, a 31% increase. That sounds small but is quite meaningful in that phones have shrunk 66%. The innards of a phone increased significantly as screens have gotten bigger, radios grown more complex, and new features like cameras have been crammed in.

It should go without saying that phones have also gotten more powerful. We looked at processor clock rates. Admittedly this is not the only metric that should be used when comparing phones but we believe correlation is sufficient to serve as a valid proxy. At the turn of the millennium the median processor clock speed on a phone was 109 MHz; today the median phone in developed markets has a clock rate of 600 MHz.

Figure 23: Median mobile phone clock rate (MHz)

0

100

200

300

400

500

600

700

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

CPU (MHz)


Another metric is memory capacity. We looked at both ROM and RAM. We see ROM as a good proxy for phone complexity as it is used largely to store the operating system and manage low-level phone operations. This has increased from 8 MB in 2002 to 512 MB for the median phone today. ROM grew at a slow, linear pace for many years and then spiked in 2008 as the iPhone came to market and smartphones became more common. ROM spiked again dramatically in 2010 which we believe is a function of the Android explosion shifting the whole mix.



Figure 24: Median mobile phone ROM capacity (MB)

0

100

200

300

400

500

600

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

ROM (MB)


RAM memory is different; it is typically used to as a storage medium for media content on a phone. There are other uses as well, and to be fair, we think the data on this subject is a little misleading. The quality of data available on the subject is a little spotty. RAM has other uses and in some cases the full capacity of the phone is not reported or skewed by multiple phone versions (e.g. 16GB and 32GB iPhone), which is not captured in our database. Setting data quality aside, the rise in RAM is also quite dramatic, and if anything we think the available data probably understates true capacity by a considerable margin. Again the pattern is clear: slow linear growth until 2007 and then the curve becomes exponential as smartphones start to flood the market.

Figure 25: Median mobile phone RAM capacity (MB)

0

50

100

150

200

250

300

350

400

450

500

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

RAM (MB)


We should also note that the data in the graphs above represent median figures. This is skewed by the still large number of feature phones on the market. We think our data captures a wide swath of all phones on the market. We will spare you the scatter plots, but in that data it is clear there are a growing number of phones that sit far above the trend line which is weighed down by all the feature phones still out there.

In short, phone capacity is now growing at rates comparable to Moore’s Law. Prior to the launch of the iPhone, we think the benefits of Moore’s Law were largely spent on bringing phone prices down. Since 2007, that trend has now been applied to phone capabilities. This is a major shift that looks set to continue for some time.



Components

We see three key areas of component advances: processors; input/output devices; and radios. Most of these businesses can rely on Moore’s Law to add capacity and features steadily. Another common theme is that most of these segments are highly competitive, but as we detail below the growing complexity of systems and greater size constraints on phones means that companies with the ability to integrate and push the R&D envelope can slowly move ahead of competitors. We have seen this in basebands, and increasingly in radio front-end modules, and believe this process will continue to winnow out smaller vendors in many fields.

Processors Smartphones have two primary silicon ‘engines’: the baseband which connects the device to the network and is something that all phones have; and another processor which handles the other computing functions, which is commonly called an Applications Processor or App Processor (AP). In many cases, this is a single chip or a single package but sourced from a single vendor. The debate about having an ‘integrated’ baseband and AP is at this stage somewhat of a religious debate pitting two camps with very strong emotions against each other with arguments often based on belief more than fact. As with most things, we think the market will ultimately reach a compromise. We believe there will always be a segment of high-performance phones which demand separate chips, but much of the market will eventually move down the integrated path because it offers better economics – one chip is usually cheaper than two.

This segment is widely covered on the Street, so we will leave it out of this note, but address it in other team publications in the future. The key vendors in the baseband space are Qualcomm, Mediatek, Intel with the recently acquired wireless arm of Infineon, ST Ericsson, Broadcom and Marvell. All of these companies sell integrated baseband and AP solutions, or at least have them on the roadmap. Discrete AP vendors include Nvidia, Freescale and Texas Instruments.

Connectivity A growing adjacency is the realm of ‘connectivity’ which refers to all the communications devices in a smartphone other than the core cellular connection. Today this includes Wi-Fi, Bluetooth, GPS and FM radios. In the near future, it will also likely include Near-Field Communications (NFC), and beyond that we could see some room for a wireless connection to HD TVs (there are several candidates vying for this) and potentially zigbee to interface with smart meters in the home (again there are several technologies vying for this slot as well, and inclusion in the smartphone at this point is science fiction).

Today the leading connectivity vendors are Broadcom, Marvell, Atheros (which Qualcomm is in the process of acquiring), CSR, Texas Instruments, Ralink (which Mediatek is in the process of acquiring), and RDA Micro. As the list above should make clear, the connectivity piece looks set to become part of the baseband. Among the other baseband vendors: ST Ericsson has announced connectivity products, and Infineon Wireless (part of Intel) has GPS but no mobile Wi-Fi or Bluetooth.

Even among baseband vendors with connectivity products, this chip tends to be standalone. As processor speeds increase, we think the connectivity features will increasingly be merged into this silicon leaving even less economic room for the stand-alone vendors. That being said, we think the combinations of possible connectivity pieces means this market is growing strongly.



Radios A related field is the radio chain of phones. Basebands produce digital signals; these need to be translated in to digital signals in radio frequencies. This is done in the transceiver and for a variety of technical reasons this work is best done in conjunction with the baseband. As a result, most transceivers today are sold by baseband vendors. Radio signals then need to be translated into analog radio waves and then sent or received by an antenna. This bit, in between the transceiver and the antenna, is now commonly known as the front-end and remains a highly competitive field. The typical front-end module contains some combination of power amplifiers (PA) filters, switches and tuners.

For radio waves to reach the base station they need to have much high power levels than signals used for digital processing in the baseband. At these levels, standard CMOS silicon does not work so well, meaning that a lot of power is consumed and lost in the translation. So, most front-ends are built using other materials, typically Gallium Arsenide (GaAs). Setting aside the acronyms, this means that it is much harder for the baseband vendors to integrate these functions into the baseband as they have done with many other functions. Working with these alternative materials is a whole separate, somewhat arcane field that is usually beyond the core competency of most of the pure silicon-base vendors. By analogy, in the automotive arena, companies that make engines and drive trains out of steel typically do not make tires out of rubber.

There are several private companies out there today looking to make front-end components out of standard silicon. This is a tricky process. Over the years, many companies have tried to do this, but very few have been able to match the performance of parts made from GaAs. Despite this, the benefits of silicon are large enough that people keep trying. To date, the only use for CMOS front-ends is in low cost handsets where performance expectations are generally low.

The GaAs landscape is comprised of seven vendors: Skyworks, RF Micro Devices, Avago, Triquint, Renasas, Anadigics and RDA Micro. All of these are public except for Renasas which is a joint venture of several Japanese electronics vendors’ semiconductor arms. While this is still a large field, it is much smaller than it was four years ago when there were 30 or so power amplifier vendors. Since then the space has gradually consolidated. Two forces drove this. First, the transition from 2G to 3G proved an expensive R&D investment which many vendors found to be not worth the effort or too far from their core competency. Second, the growing complexity meant that a better path was to integrate all the PAs into a single module. Companies that could do this integration were able to steal a march on those who could not. And the handset vendors who make these purchasing decisions preferred to work with integrated solutions rather than buy individually and do the integration themselves.

We think the rise of smartphones and the coming transition to 4G will likely further narrow the field for all the same reasons we saw in the 2G to 3G transition. One of the difficulties facing smartphone makers is the need to encompass all the radio bands used in the market. The typical 2G phone has four or five bands, 3G phones use three or four. For 4G phones, the situation is even worse with carriers looking to deploy LTE in seven different bands around the globe. Phones will not need to support all of these, but they will need to support a lot of them, and this will make for some very complicated front-ends.

The PA and front-end makers are responding to this by bundling multiple PAs into a single chip. For years, front-end modules tended to have one PA for each band. Increasingly, the industry is moving to multi-mode or multi-band technologies. In this scheme a single PA will handle lower bands (say 700 MHz – 900 MHz) and a second PA will handle the higher bands (1800 MHz – 2600 MHz). Integrating these bands is again a difficult R&D problem and comes with further problems in the need for very high performance switching and filtering. All of this favors the largest vendors such as Skyworks.



We think the transition to smartphones will prove to be a big boost for the companies which survive in the field. First, as noted above, they will likely see fewer competitors which should help pricing and expand the available share. Secondly, this growing complexity will mean more front-end content per phone helping their revenue. In general, we think Skyworks is the best positioned here. RFMD appears challenged by several years of underinvestment in R&D and strategic missteps. That being said, they have several major customers and are widely regarded as having the best engineers in the space, if not the best management team. Triquint appears well positioned technically, but struggles to match the scale of larger vendors. Avago continues to execute well, but does not have all the parts needed for a complete solution. Anadigics continues to struggle with its size and customer line-up. Renesas supplies almost solely to Nokia and does not look to be well positioned for the transition to 4G. RDA Micro looks firmly entrenched selling to the China handset ecosystem.

Input/Output One of the most interesting phenomena of smartphones is the extent to which they are being used for a whole new range of tasks unimaginable a few years ago. In particular, smartphones are being used as a new form of sensory devices. For instance, social networking sites now see a huge amount of traffic from people uploading photos combined with GPS data. This is not only updates from family vacations but a range of other tasks like Foursquare check-ins and social gaming. Privately-held Smule has made the iPhone into a range of musical instruments. We have lost track of how many start-ups are using smartphone cameras as barcode scanners. Our favorite example, we have heard several companies may use phone cameras to take a picture of a user’s fingerprints, and use that as a form of authentication in lieu of a password. Those are just the input devices. On the output side, the case for better screens and microphones is straightforward – users will always want better performance.

We think application developers will continue to find news to make use of phone input/output devices and as a result the need for good components will grow. This segment covers a wide swath of suppliers many of whom are based in Asia. We will not cover all of them here, but believe this area has meaningful potential.

Categories to consider:

Camera sensors and camera modules: Camera sensors and camera modules: There seems to be a perpetual race in this space as handset vendors bring increasingly high-resolution cameras to the market. For instance, some recent blogosphere analysis (e.g., blog.geekaphone.com) holds that the iPhone has become the world’s most digital camera.

That being said, in our view, beyond a certain point having more megapixels becomes pointless. Take too many 12 MP photos and they can quickly eat up all the storage space on a phone and become very difficult to manage, upload and share. At some point, megapixel resolution will become less of a differentiator. However, the industry is showing new forms of innovation, in the form of cheaper and thinner camera modules.

Audio: Speakers and microphones and voice quality are increasingly common topics of conversation. The smartphone industry has been heavily influenced by Apple and the iPhone carries heavy influences from the iPod music player. As a result, music is an area that all the smartphone vendors have to contend with. So far, no one has come up with a software/hardware combination to rival Apple’s, but that has not stopped them from trying. We also think music playback remains hugely important to consumers, and in markets such as India this includes both personal playback in earphones as well as sharing the audio through speakers. As a result, we think audio quality will grow in importance.



We think two areas in this space merit attention, albeit for different reasons. First is analog audio components. Here we would look at Hong Kong-listed AAC Acoustic. AAC makes a wide range of speakers, microphones and other audio products (as well as other input/output components). As a vendor to many leading phone platforms, we think they are well positioned to see further gains. A less publicized area is speech processing on the handset. For instance, privately-held Audience offers a voice processing chip which looks to improve speech quality of cellular connections through enhanced noise cancellation and other improvements. This is one way which we believe not only handset vendors, but carriers as well, may seek to differentiate their offerings.

Screens: Yet one more area wherein Apple is driving the industry. With the launch of the iPhone4 they made screen resolution a marketing tool so now other companies are looking to increase pixel density in their phones. The LCD screen market is a complex industry in its own right, beyond our scope here. Nonetheless, we see this as an important growth area as companies compete with higher pixel densities and new screen materials such as AMOLED from Samsung, and eventually Qualcomm’s Mirasol or privately held Pixel Qi’s transmissive ‘dual mode’ screens.



Future features As in past reports, we like to end with a somewhat more open-ended look at topics that may not fit easily into the rest of the report, but which we feel could be important or at the very least interesting enough to merit further attention in upcoming reports.

Here we look at several:

Feature phones

In this report we have focused on smartphones. This is the more dynamic piece of the industry, but it is important to bear in mind that smartphones are only a portion of the total handset market. By our estimates, smartphones will contribute only about 25% of total handset units this year. The installed base for smartphones is about 300 million today, while the installed base for all handsets is about 5.5 billion. Even with the explosion in low-cost smartphones that we expect, we see the installed base only reaching 1 billion by 2015, at which point global mobile phone penetration will likely exceed six or seven billion, that is to say every man, woman, child and robot on the planet will on average have at least one phone. So ignoring feature phones is a mistake for developers with global ambitions.

One of the companies that appears to be most aware of this is Facebook. They recently rolled a unified mobile site that scales according to the capability of the device used to access their site. This is a difficult task, something that has stymied developers for years. We point it out for this reason. Before the advent of mobile OS, every application had to be rewritten or ported over to multiple phone platforms. Game companies, for instance, would have to create 1,500 different versions of application code bases to support all the devices requested by carriers. The mobile OS made this process easier, and the mobile web has made it almost seamless. An important part of this is 0.Facebook. This is a stripped down version of their mobile site which has no photos and carries a very light load on the network. This is meant for data-ready 2G feature phones and typically requires only a very limited data plan. The company has even taken this a step further by working with Gemalto to embed Facebook features on SIM cards. Users with this feature can post to their Facebook page and get updates using SMS text messages. This essentially puts Facebook within reach of the 80% of mobile phone users, even those using some of the most basic phones out there.

Facebook can do this because of their prominence which helps them make deals with carriers. However, we would argue that other developers should look for ways to go after the feature phone market. This is particularly true for websites with social networking features. For those companies, the size of the audience matters immensely; ignoring the vast 2G masses could prove to be a mistake. That said, we recognize there are many difficulties in tackling such a strategy. Few startups have the resources to court carrier deals, and as noted above, this has proved a painful process in the past. We think this creates an opportunity for some other platform to emerge that addresses this need. Such a solution could be accomplished via a partnership with the handset makers or baseband vendors such as Qualcomm or Mediatek who have a wide reach.

Mobile payments

We have written extensively on this topic in recent months (let us know if you would like a copy of past reports including our primer from last October). We think this year will mark the start of a major adoption of the technologies needed to enable mobile payments in smartphones. In particular, we think NFC chips will become a standard feature in smartphones. We think 50 million phones will include this feature this year, a number which could swing dramatically depending on how Nokia and Apple plan to get more aggressive in the space.



Mobile payments is an area which has held promise for years, but has proved difficult to achieve. This difficulty stems from the fact that an entire ecosystem of major institutions need to agree on the terms. These institutions include banks, wireless carriers, payment processors and handset vendors. Getting any of these industries to agree has been difficult. This is starting to change as the technology advances. First and foremost is the inclusion of NFC in a variety of handsets. This now seems to be proceeding regardless of the mobile payments systems. Content vendors, OS providers and smartphone makers have found reasons to include NFC that are unrelated to mobile payments. We see this is paving the way once the payment side catches up. Moreover, several players – notably Apple and Google – appear to be working on payment systems that have sufficient critical mass to exist without full-blown institutional support. As these become more predominant, we think the other players will accelerate their plans or risk being left behind.

In mobile payments, the industry has looked enviously at Japan’s payment system for many years. The carriers there launched a mobile payment almost ten years ago that is now widely deployed. Mobile consumers can use their phones to make purchases at most retail locations vending machines, and buy tickets for public transit systems. However, Japanese carriers had a major advantage. When they first launched their systems, credit card penetration in Japan was very low. Their mobile payment systems thus emerged in a payment vacuum. Consumers were happy to switch to mobile payments because the alternative was cash. In the US and Europe, credit and debit cards are already widely deployed so there is less obvious demand from consumers. In these markets, we think the non-payment uses for NFC will likely prove to be more important in boosting consumer adoption. In past reports, we noted uses of NFC as varied as hotel Wi-Fi configuration and corporate key card access to offices. As this ecosystem progresses, we think consumers will use these applications as their introduction to NFC and with time learn about the payment systems.

On the mobile side of this ecosystem, we think the key dynamic to watch will be the location of the ‘secure element’ in phones. The secure element is essentially a piece of software that has a level of cryptographic processes which ensure that the device is associated with a specific owner. The wireless carriers would like to see the secure element embedded in the SIM cards they provide. This would mean that all payment systems would have to dip into the carriers’ subscriber database to authenticate that the payment user is in fact tied to a real identity. In doing so, the carriers would give themselves an important role in authenticating mobile payments. Alternatively, handset vendors and OS platform providers would like to see the secure element sit in a standalone chip or some piece of software. This would funnel authentication requests through some other payment database such as Google Checkout or Apple’s iTunes Store. Credit card processors and banks are working with partners on both sides of this divide as well as experimenting with other alternatives.

Currently, the leading providers of NFC chip technology are NXP and Inside Secure. Broadcom also recently acquired a company to gain access to this technology. As noted in the connectivity section in this report, we think the other baseband vendors will soon enter this market and eventually embed the technology in the apps processor or the baseband. On the equipment side, Verifone, Ingenico and VivoTech are emerging as the leaders in providing the merchant payment systems that would serve as NFC readers. We would also add Gemalto as an important partner in the mobile payment space. They are a provider of SIM cards which are one of the major security and crypto elements in many mobile payment schemes.

Augmented Reality

Another interesting field is Augmented Reality. This grand term refers to using smartphones to layer on additional data to real world images. For instance, we have seen demos where users hold up smartphones of a street scene. On the phone monitor, the camera displays an



image of the street with additional data layered on top. This data could include driving directions, descriptions of the buildings on the screen, or data from social networks showing locations of nearby friends.

Developers are in early stages of experimenting with what could become a very powerful tool. We see this as not only important to developers, but the industry as a whole. Firstly, content providers could make use of these ‘AR’ systems to make their services much richer and thus more valuable. It is one thing to get a text message about a nearby deal or coupon, quite another to hold up your phone and see what the local merchants are offering. Secondly, as these systems become more powerful we think they could substantially add to data traffic. Today, carriers are planning for a world in which users are constantly downloading some amount of data – voice, video, text. AR systems will draw down many more layers later and multiply those video and image downloads. It will no longer be enough to upload a video stream of the local surroundings; phones will also have to download position data, merchant data, social network data, and several other layers. The science fiction images of Augmented Reality scenarios are notorious for being very busy, and we think as the technology edges closer to maturity it will add a layer of network loading that no one is currently expecting.

M2M – Skynet’s First Steps

Another area that always seems to be “just a few years away” is Machine to Machine Communication. This is commonly known as M2M, and the fact that it has that ‘2’ in the middle should serve as an indicator for how long the idea has been out there.

The idea behind M2M is that some remote computer or sensor or other piece of electronics can be placed ‘out there’, somewhere in the world and then communicate over the cellular network with the Internet or the Cloud or some network. This technology is slowly emerging, and is most familiar to people in the form of Smart Meters, which are likely to become the first scale M2M deployments. The industry and the blogosphere are full of interesting uses for M2M technologies. From the ‘connected refrigerator’ that sends you a text message when you are low on milk to vast sensor networks that record temperature readings on every container on Panamax freight ship.

The area of M2M merits a FITT report in its own right, but we think developments in the smartphone OS space have important implications for the field.

One of the big challenges of M2M is that today all of these networks are essentially built on proprietary communications platforms. Every network is essentially custom-made. This is expensive and presents a huge barrier. We hold the view that for M2M to truly take off someone will have to build platforms to run them on. There are several pieces to this, but one of the key elements will be software drivers that can communicate with the network. Today, every prospective M2M player has to write their own telephony stack. In the next year or two, we expect to see increasing use of Android to power Android modules. Strictly speaking, this is not necessary as Android has far more features than a typical M2M device will need. However, we believe Android can actually operate on a relatively small processor and memory footprint. Any added cost needed to support Android on the device would be tiny compared with the benefit of not having to rewrite that telephony stack. M2M makers could take off-the-shelf Android, write a driver for the particular device they are building (a relatively easy task). We see Android as removing a major pain point in deploying M2M networks. We have heard that several companies including a network operator may be looking at just this approach.



The nature of phones

Finally, we want to indulge in a bit of abstract thinking about the nature of phones. Today, phones are seen as discrete devices. People have a phone; they have a computer; they have an iPad. But all of these are starting to resemble each other. The iPhone and the iPad are the best example: except for the screen, the internal components are essentially identical with each other. As processing power increases, the distinction between these devices essentially goes away, and the only difference becomes input/output methods – this is just a matter of screen size and keyboards…

We think that for years to come, all of these form factors will continue to coexist, but at heart there is no reason for them to be separate. With time, small devices, which we call phones today, will have a full array of output methods. For instance, most Android phones today have HDMI ports. As this process evolves, users will continue to carry the phones in their pockets; tablets and laptops will never be portable. Instead, we will start to use the small device as our main computing device. When you want to do more intensive typing or watch content on a larger phone, you will connect those to the phone. We are starting to see the reemergence of the ‘docking station’ with Motorola’s weptop accessory for the Atrix. With time, the docking station will give way to wireless connection (via Bluetooth, Wi-Fi or wireless HD of some sort) to the Input and Output device of your choice. People will not need separate computers; they will be able to connect their smartphone to a bigger screen and keyboard, and this will become the computer. The MiFi, wireless hotspot, is another step in this direction. And this is not only a consumer trend; it will work in the enterprise as well. RIM, Qualcomm, Google and many others are already working on ways to virtualize phones to provide personal and enterprise profiles on the phone.

This puts the value of the ‘phone’ into two key areas. Firstly, it is a ubiquitous connection to the network, which means that modems, connectivity and basebands will remain highly valuable. Secondly, and we believe more importantly, the small device will become a form of identity. Phones will come with permissions to work on certain networks and access certain data based on the owner’s profile. This will serve to connect to large content such as videos and corporate systems and to connect to small content such as payment streams, and everything in between. The key piece will be the identity of the phone. Communications will be widespread, but identity will be the scarce commodity.

Deutsche B

ank Securities Inc.

Page 63

11 April 2011

Wireless Equipm

ent M

obile Operating S

ystems

Appendix A: Cellular standards roadmap Figure 26: Cellular standards roadmap

802.16d

802.11n& Mesh

802.16e

UMTS-TDD(TD-CDMA)

TD-SCDMA

CDMA (IS-95B)

GSM/GPRSGSM

PDC

TDMA

CDMA(IS-95A

iDEN WiDEN HSDPA

1xEV-DO

WCDMA(UMTS-FDD)

CDMA2000

EDGE

1xEV-DOa

HSUPA HSPA+

1xEV-DOb

802.20(OFDM)

LTE




Appendix 1 Important Disclosures

Additional information available upon request

Disclosure checklist Company Ticker Recent price* Disclosure Skyworks Solutions SWKS.OQ 27.55 (USD) 8 Apr 11 2,6 Atheros Communications ATHR.OQ 44.65 (USD) 8 Apr 11 2 Motorola Mobility MMI.N 24.03 (USD) 8 Apr 11 8 Qualcomm QCOM.OQ 53.63 (USD) 8 Apr 11 2,6,8,14,15 Research In Motion RIMM.OQ 54.78 (USD) 8 Apr 11 2 *Prices are sourced from local exchanges via Reuters, Bloomberg and other vendors. Data is sourced from Deutsche Bank and subject companies.

Important Disclosures Required by U.S. Regulators Disclosures marked with an asterisk may also be required by at least one jurisdiction in addition to the United States. See “Important Disclosures Required by Non-US Regulators” and Explanatory Notes. 2. Deutsche Bank and/or its affiliate(s) makes a market in securities issued by this company.

6. Deutsche Bank and/or its affiliate(s) owns one percent or more of any class of common equity securities of this company calculated under computational methods required by US law.

8. Deutsche Bank and/or its affiliate(s) expects to receive, or intends to seek, compensation for investment banking services from this company in the next three months.

14. Deutsche Bank and/or its affiliate(s) has received non-investment banking related compensation from this company within the past year.

15. This company has been a client of Deutsche Bank Securities Inc. within the past year, during which time it received non-investment banking securities-related services.

Important Disclosures Required by Non-U.S. Regulators Please also refer to disclosures in the “Important Disclosures Required by US Regulators” and the Explanatory Notes. 2. Deutsche Bank and/or its affiliate(s) makes a market in securities issued by this company.

6. Deutsche Bank and/or its affiliate(s) owns one percent or more of any class of common equity securities of this company calculated under computational methods required by US law.

For disclosures pertaining to recommendations or estimates made on securities other than the primary subject of this research, please see the most recently published company report or visit our global disclosure look-up page on our website at http://gm.db.com/ger/disclosure/DisclosureDirectory.eqsr.

Analyst Certification

The views expressed in this report accurately reflect the personal views of the undersigned lead analyst about the subject issuers and the securities of those issuers. In addition, the undersigned lead analyst has not and will not receive any compensation for providing a specific recommendation or view in this report. Jonathan Goldberg



Historical recommendations and target price: Skyworks Solutions (SWKS.OQ)

(as of 4/8/2011)

1

23 4

56

7 89

101112

13

1415

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

Apr 08 Jul 08 Oct 08 Jan 09 Apr 09 Jul 09 Oct 09 Jan 10 Apr 10 Jul 10 Oct 10 Jan 11

Se

curit

y Pr

ice

Da te

Previous Recommendations

Strong Buy Buy Market Perform Underperform Not Rated Suspended Rating

Current Recommendations

Buy Hold Sell Not Rated Suspended Rating

*New Recommendation Structure as of September 9, 2002

1. 7/18/2008: Hold, Target Price Change USD9.00







8. 2/22/2010: Upgrade to Buy, Target Price Change USD18.00

9. 4/30/2010: Buy, Target Price Change USD20.00









Historical recommendations and target price: Atheros Communications (ATHR.OQ)

(as of 4/8/2011)

1

2

3

4 5

6

7 8

9

10

1112

13

14

15

16

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

45.00

50.00


Se

curit

y Pr

ice

Da te










5. 12/18/2008: Downgrade to Hold, Target Price Change USD14.00



8. 6/4/2009: Upgrade to Buy, Target Price Change USD25.00








16. 1/18/2011: Downgrade to Hold, Target Price Change USD45.00

Historical recommendations and target price: Motorola Mobility (MMI.N)

(as of 4/8/2011)

1

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

Dec 10 Mar 11

Se

curit

y Pr

ice

Da te









Historical recommendations and target price: Qualcomm (QCOM.OQ)

(as of 4/8/2011)

1

2

34

56

7 8

9

10

11

12

13

14

15

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00


Se

curit

y Pr

ice

Da te





















Historical recommendations and target price: Research In Motion (RIMM.OQ)

(as of 4/8/2011)

1

2

34

5

67

89

10

1112

0.00

20.00

40.00

60.00

80.00

100.00

120.00

140.00

160.00


Se

curit

y Pr

ice

Da te






1. 9/26/2008: Downgrade to Sell, Target Price Change USD70.00

2. 10/3/2008: Sell, Target Price Change USD50.00



5. 4/3/2009: Upgrade to Hold, Target Price Change USD56.00



8. 12/18/2009: Upgrade to Hold, Target Price Change USD75.00







Equity rating key Equity rating dispersion and banking relationships

Buy: Based on a current 12- month view of total share-holder return (TSR = percentage change in share price from current price to projected target price plus pro-jected dividend yield ) , we recommend that investors buy the stock.

Sell: Based on a current 12-month view of total share-holder return, we recommend that investors sell the stock

Hold: We take a neutral view on the stock 12-months out and, based on this time horizon, do not recommend either a Buy or Sell.

Notes: 1. Newly issued research recommendations and target prices always supersede previously published research.

2. Ratings definitions prior to 27 January, 2007 were:

Buy: Expected total return (including dividends) of 10% or more over a 12-month period

Hold: Expected total return (including dividends) between -10% and 10% over a 12-month period

Sell: Expected total return (including dividends) of -10% or worse over a 12-month period

48 % 50 %

2 %

42 % 34 %

31 %0

50100150200250300350400450500

Buy H old S ell

N orth Am e rican U n iv e rse

C om pan ie s C o vere d C o s. w/ B an k in g R e latio n ship



Regulatory Disclosures

1. Important Additional Conflict Disclosures

Aside from within this report, important conflict disclosures can also be found at https://gm.db.com/equities under the "Disclosures Lookup" and "Legal" tabs. Investors are strongly encouraged to review this information before investing.

2. Short-Term Trade Ideas

Deutsche Bank equity research analysts sometimes have shorter-term trade ideas (known as SOLAR ideas) that are consistent or inconsistent with Deutsche Bank's existing longer term ratings. These trade ideas can be found at the SOLAR link at http://gm.db.com.

3. Country-Specific Disclosures

Australia: This research, and any access to it, is intended only for "wholesale clients" within the meaning of the Australian Corporations Act. Brazil: The views expressed above accurately reflect personal views of the authors about the subject company(ies) and its(their) securities, including in relation to Deutsche Bank. The compensation of the equity research analyst(s) is indirectly affected by revenues deriving from the business and financial transactions of Deutsche Bank. EU countries: Disclosures relating to our obligations under MiFiD can be found at http://globalmarkets.db.com/riskdisclosures. Japan: Disclosures under the Financial Instruments and Exchange Law: Company name - Deutsche Securities Inc. Registration number - Registered as a financial instruments dealer by the Head of the Kanto Local Finance Bureau (Kinsho) No. 117. Member of associations: JSDA, The Financial Futures Association of Japan. Commissions and risks involved in stock transactions - for stock transactions, we charge stock commissions and consumption tax by multiplying the transaction amount by the commission rate agreed with each customer. Stock transactions can lead to losses as a result of share price fluctuations and other factors. Transactions in foreign stocks can lead to additional losses stemming from foreign exchange fluctuations. "Moody's", "Standard & Poor's", and "Fitch" mentioned in this report are not registered credit rating agencies in Japan unless “Japan” is specifically designated in the name of the entity. New Zealand: This research is not intended for, and should not be given to, "members of the public" within the meaning of the New Zealand Securities Market Act 1988. Russia: This information, interpretation and opinions submitted herein are not in the context of, and do not constitute, any appraisal or evaluation activity requiring a license in the Russian Federation.


North American locations

Deutsche Bank Securities Inc. 60 Wall Street New York, NY 10005 Tel: (212) 250 2500

Deutsche Bank Securities Inc. One International Place 12th Floor Boston, MA 02110 United States of America Tel: (1) 617 217 6100

Deutsche Bank Securities Inc. 222 South Riverside Plaza 30th Floor Chicago, IL 60606 Tel: (312) 537-3758

Deutsche Bank Securities Inc. 3033 East First Avenue Suite 303, Third Floor Denver, CO 80206 Tel: (303) 394 6800

Deutsche Bank Securities Inc. 1735 Market Street 24th Floor Philadelphia, PA 19103 Tel: (215) 854 1546

Deutsche Bank Securities Inc. 101 California Street 46th Floor San Francisco, CA 94111 Tel: (415) 617 2800

Deutsche Bank Securities Inc. 700 Louisiana Street Houston, TX 77002 Tel: (832) 239-4600

International locations

Deutsche Bank Securities Inc. 60 Wall Street New York, NY 10005 United States of America Tel: (1) 212 250 2500

Deutsche Bank AG London 1 Great Winchester Street London EC2N 2EQ United Kingdom Tel: (44) 20 7545 8000

Deutsche Bank AG Große Gallusstraße 10-14 60272 Frankfurt am Main Germany Tel: (49) 69 910 00

Deutsche Bank AG Deutsche Bank Place Level 16 Corner of Hunter & Phillip Streets Sydney, NSW 2000 Australia Tel: (61) 2 8258 1234

Deutsche Bank AG Filiale Hongkong International Commerce Centre, 1 Austin Road West,Kowloon, Hong Kong Tel: (852) 2203 8888

Deutsche Securities Inc. 2-11-1 Nagatacho Sanno Park Tower Chiyoda-ku, Tokyo 100-6171 Japan Tel: (81) 3 5156 6770

Global Disclaimer The information and opinions in this report were prepared by Deutsche Bank AG or one of its affiliates (collectively "Deutsche Bank"). The information herein is believed to be reliable and has been obtained from public sources believed to be reliable. Deutsche Bank makes no representation as to the accuracy or completeness of such information.

Deutsche Bank may engage in securities transactions, on a proprietary basis or otherwise, in a manner inconsistent with the view taken in this research report. In addition, others within Deutsche Bank, including strategists and sales staff, may take a view that is inconsistent with that taken in this research report.

Opinions, estimates and projections in this report constitute the current judgement of the author as of the date of this report. They do not necessarily reflect the opinions of Deutsche Bank and are subject to change without notice. Deutsche Bank has no obligation to update, modify or amend this report or to otherwise notify a recipient thereof in the event that any opinion, forecast or estimate set forth herein, changes or subsequently becomes inaccurate. Prices and availability of financial instruments are subject to change without notice. This report is provided for informational purposes only. It is not an offer or a solicitation of an offer to buy or sell any financial instruments or to participate in any particular trading strategy. Target prices are inherently imprecise and a product of the analyst judgement.

As a result of Deutsche Bank’s recent acquisition of BHF-Bank AG, a security may be covered by more than one analyst within the Deutsche Bank group. Each of these analysts may use differing methodologies to value the security; as a result, the recommendations may differ and the price targets and estimates of each may vary widely.

Deutsche Bank has instituted a new policy whereby analysts may choose not to set or maintain a target price of certain issuers under coverage with a Hold rating. In particular, this will typically occur for "Hold" rated stocks having a market cap smaller than most other companies in its sector or region. We believe that such policy will allow us to make best use of our resources. Please visit our website at http://gm.db.com to determine the target price of any stock.

The financial instruments discussed in this report may not be suitable for all investors and investors must make their own informed investment decisions. Stock transactions can lead to losses as a result of price fluctuations and other factors. If a financial instrument is denominated in a currency other than an investor's currency, a change in exchange rates may adversely affect the investment. Past performance is not necessarily indicative of future results. Deutsche Bank may with respect to securities covered by this report, sell to or buy from customers on a principal basis, and consider this report in deciding to trade on a proprietary basis.

Unless governing law provides otherwise, all transactions should be executed through the Deutsche Bank entity in the investor's home jurisdiction. In the U.S. this report is approved and/or distributed by Deutsche Bank Securities Inc., a member of the NYSE, the NASD, NFA and SIPC. In Germany this report is approved and/or communicated by Deutsche Bank AG Frankfurt authorized by the BaFin. In the United Kingdom this report is approved and/or communicated by Deutsche Bank AG London, a member of the London Stock Exchange and regulated by the Financial Services Authority for the conduct of investment business in the UK and authorized by the BaFin. This report is distributed in Hong Kong by Deutsche Bank AG, Hong Kong Branch, in Korea by Deutsche Securities Korea Co. This report is distributed in Singapore by Deutsche Bank AG, Singapore Branch, and recipients in Singapore of this report are to contact Deutsche Bank AG, Singapore Branch in respect of any matters arising from, or in connection with, this report. Where this report is issued or promulgated in Singapore to a person who is not an accredited investor, expert investor or institutional investor (as defined in the applicable Singapore laws and regulations), Deutsche Bank AG, Singapore Branch accepts legal responsibility to such person for the contents of this report. In Japan this report is approved and/or distributed by Deutsche Securities Inc. The information contained in this report does not constitute the provision of investment advice. In Australia, retail clients should obtain a copy of a Product Disclosure Statement (PDS) relating to any financial product referred to in this report and consider the PDS before making any decision about whether to acquire the product. Deutsche Bank AG Johannesburg is incorporated in the Federal Republic of Germany (Branch Register Number in South Africa: 1998/003298/10). Additional information relative to securities, other financial products or issuers discussed in this report is available upon request. This report may not be reproduced, distributed or published by any person for any purpose without Deutsche Bank's prior written consent. Please cite source when quoting.

Copyright © 2011 Deutsche Bank AG

mobile os overview april 2011

Education

future landscape

mobile thought

smartphone os landscape

smartphone market

pc landscape

onlyfor mobile

smartphone share

research reports