2015 global field programmable gate array (fpga ... · gate array (fpga) product lines; these are...

2015

2015 Global Field Programmable Gate Array (FPGA) Technology Innovation Leadership Award

BEST PRACTICES RESEARCH

© Frost & Sullivan 2015 2 “We Accelerate Growth”

Contents

Background and Company Performance ........................................................................ 3

Industry Challenges .............................................................................................. 3

Technology Attributes and Future Business Value ..................................................... 3

Conclusion........................................................................................................... 7

Significance of Technology Innovation .......................................................................... 8

Understanding Technology Innovation .......................................................................... 8

Key Benchmarking Criteria .................................................................................... 9

Best Practice Award Analysis for Altera ......................................................................... 9

Decision Support Scorecard ................................................................................... 9

Technology Attributes ......................................................................................... 10

Future Business Value ......................................................................................... 10

Decision Support Matrix ...................................................................................... 11

The Intersection between 360-Degree Research and Best Practices Awards ..................... 12

Research Methodology ........................................................................................ 12

Best Practices Recognition: 10 Steps to Researching, Identifying, and Recognizing Best Practices ................................................................................................................. 13

About Frost & Sullivan .............................................................................................. 14



Background and Company Performance

Industry Challenges

With the exponential increase in data traffic, facilitated by the increasing adoption of

mobility and the Internet of Things (IoT), companies are struggling to better manage data

center workloads while accelerating performance. Aware of such growing concerns,

contemporary semiconductor companies have sought to deliver robust field programmable

gate array (FPGA) product lines; these are programmable semiconductors that can offload

the CPU in a server and accelerate computing to keep up with the influx of Big Data and

support myriad new applications running in the data center. Although the market has

several vendors offering FPGAs who are claiming excellent performance, until recently

FPGAS utilized a fixed point architecture, which is incapable of representing very large, or

very small numbers. To address customer pain points and deliver excellent performance in

an FPGA, it has become important for FPGAs to have digital signal processors (DSPs) with

native hard floating point computational units. Additionally, there has been significant

evolution in programming languages that play to this requirement. While traditional fixed

point FPGAs can support Very High Speed Integrated Circuit (VHSIC), Hardware

Description Language (VHDL), and VeriLog, programmers now mostly use high-level

programming languages such as C, MATLAB, and Simulink. Frost & Sullivan points out that

FPGAs need floating point capabilities to leverage the productivity offered by these

languages.

However, Frost & Sullivan independent analysis confirms that delivering such an advanced

solution has its own set of challenges. While many semiconductor companies have

attempted to implement floating point operators into FPGAs, they report to Frost &

Sullivan that they have not been successful in delivering efficient performance, reducing

power consumption, and increasing speed at the same time. In addition, FPGAs come with

thousands of computational units, which makes the incorporation of floating point

operators even more challenging. To engineer such a futuristic solution, companies require

immense skill and expertise and must make a significant investment in research and

development (R&D).

In this context, Frost & Sullivan expects FPGA vendors who address the customer pain

points by developing best-in-class floating point FPGAs to enjoy a leading position in the

global FPGA market.

Technology Attributes and Future Business Value

Industry Impact

Founded in 1983, and based in San Jose, California, Altera Corporation (Altera), is a global

leader in manufacturing programmable logic devices (PLDs). Altera revolutionized digital

signal processing (DSP) with its groundbreaking introduction of IEEE 754 single precision



floating point in its Arria 10 FPGAs in 2014. Sharply deviating from the conventional trend

of using fixed point FPGAs in electronic systems designs for data centers, radar

technology, antenna processing, and medical imaging, the next-generation FPGAs from

Altera complement central processing units (CPUs) with thousands of hardened floating

point blocks, providing a processing rate of up to 1.5 Tera floating point operations per

second (TFLOPs), greater energy efficiency, and improved productivity.

Frost & Sullivan recognizes Altera’s innovation in the company’s commitment to enabling

its customers to overcome the limitations of fixed point FPGAs in representing numerical

systems of large and small magnitudes. The company’s ability to seamlessly integrate

both IEEE 754 single precision multiplier and adder computational units with the familiar

fixed point nodes has helped it address the industry demand for greater precision in data

processing. While some of the competing DSP systems and graphics processing units

(GPUs) provide support for floating point calculation, Altera’s Generation 10 FPGAs

(Arria® 10, and the upcoming Stratix® 10) deliver this capability along with the

ubiquitous connectivity and hardware flexibility available with FPGAs.

Compared to traditional GPUs, also known as general-purpose graphics processing units

(GPGPUs), supporting floating point computation, the DSP blocks in Altera’s Generation 10

FPGAs make these units more consistent in terms of productivity levels. With its pipelined-

logic architecture instead of a GPU-parallel-processor architecture, delivering better

performance per watt GigaFLOPS (GFLOPS/W), and enabling flexible connectivity support

for the datapath, Altera’s FPGAs provide customers with reduced latency in data

processing enhanced computational capability. The FPGAs offer increased performance,

and still provide with lowest system power consumption, and performance, while providing

the high levels of fixed point performance which FPGAs are well known for.

Product Impact

Altera’s Generation 10 FPGAs were engineered to enable these hard floating point

calculations, and has enabled the company to enter new markets. At the same time it has

enhanced the performance of its Arria 10 product line and the follow-on product line of

Stratix 10 FPGAs and SoCs, which is to be launched in the first quarter of 2016.

The soon-to-be-launched Stratix 10 FPGAs from Altera’s Generation 10 solutions

demonstrates world-class capabilities. Besides a revolutionary 14 nm device fabrication

from Intel, the product will also offer a 2x core performance enhancement over other

high-end PLDs, by attaining a speed of 1 GHz. Stratix 10 will contain Altera’s proprietary

HyperFlex™ technology which also provides the company with a competitive advantage.

While conventional architectures are often limited by issues such as bus widths, congested

routing, and interconnection difficulties, Altera’s HyperFlex™ architecture overcomes these

shortcomings and guarantees optimal performance. In addition, in conjunction with the

64-bit quad core ARM Cortex™-A53 processor, Stratix 10 promises to deliver a processing



speed of up to 10 TFLOPS, which is unmatched in the industry, along with 70 percent less

power consumption than standard FPGAs.

Visionary Innovation

Frost & Sullivan feels that one of the key strengths of Altera setting it apart from its

competitors in the global FPGA market is its strong technological expertise. With more

than three decades of experience in the industry, the company has displayed dedicated

focus and persistence in engineering industry-leading FPGAs with advanced capabilities.

Engineering in the floating point technology has empowered customers by giving them

better digital signal processing capabilities, as they now can implement their algorithms in

floating point. At the core of the solution lies Altera’s advanced variable precision DSP

blocks, each of which includes a single-precision adder and a single-precision multiplier.

The 1:1 ratio of multipliers and adders in each DSP block facilitates the independent usage

of multiply, adder, mult-add or mult-accumulator. With thousands of floating point

operators being implemented in each of these DSP blocks, Altera’s 20 nm Arria 10 range

of FPGAs has rated from 140 GFLOPS to 1.5 TFLOPS. The company’s more advanced 14

nm Stratix 10 range of FPGAs provides a peak computational rate of up to 10 TFLOPS,

which is the highest recorded speed for any semiconductor device. Further, the floating

point computational units (multiplier and adder) are integrated with the existing variable

precision fixed point modes. Thus, while designers can benefit from the fixed point DSP

processing features available in the current designs, they can also upgrade a part or all of

the design to single precision floating point for greater numerical precision, and a dynamic

range. To further boost performance, the DSP blocks support special vector modes to

leverage linear algebraic functions ideal for both high-performance computing applications

and more traditional FPGA applications, such as highly parallel fast Fourier transform

(FFT), or finite impulse response (FIR) filter implementations. Such distinct and flexible

architecture allows customers to derive the maximum benefit from floating point operators

and achieve a higher speed of up to 10 TFLOPS. With more than 70 math.h library

functions being compatible with OpenCL 1.2, the FPGAs ensure robust performance, even

in fully packed designs. Unlike competing FPGA products, there is no need to use

programmable logic for floating point computations. Also, similar clock rates as those used

in fixed point designs can support Altera’s floating point operators, even with 100% usage

of the DSP blocks.

In addition to its groundbreaking innovation of incorporating floating point operators in

FPGAs, Altera has further improvised on the architecture of the Stratix 10 range to ensure

2X core performance improvements compared to traditional high-performance

programmable devices. The company’s cutting-edge HyperFlex architecture combined with

the Intel 14 nm Tri-gate process enables this world-class performance. The core fabric

architecture, deployed in the Stratix 10 FPGA, addresses some primary issues, such as

increasing bus widths, routing congestions, and interconnection problems, delivering a

higher performance level and greater power efficiency compared to traditional fixed point



FPGAs. The new architecture ensures a floating point clock rate of about 800MHz, with

even greater clock rates in fixed point, up to 1 GHz.

Application Diversity

Altera’s impressive portfolio of floating point FPGAs demonstrates the capability to be

deployed across varied applications. With the exponential increase in data traffic owing to

the increasing adoption of mobility, video traffic, and the Internet of Things (IoT),

companies now require large-scale data centers with efficient servers and improved

storage systems. Altera’s incorporation of floating point operators in its FPGAs enables

such capabilities. With the company’s floating point DSPs, these FPGAs provide customers

with greater data processing speed, low latency, and reduced power consumption. With

the current availability of its Arria 10 product, the company expects data centers to be the

growth engine for its floating point FPGAs. The company’s innovations also can also be

applied in radar systems, medical imaging, and wireless systems. For instance, advanced

wireless systems, incorporating MIMO, military radar, CT scan instruments, and magnetic

resonance imaging (MRI), have multiple antennas absorbing huge amounts of data. Such

systems require increased dynamic range in signal processing algorithms, which is only

possible with floating point FPGAs. While Altera’s floating point FPGAs have been

increasingly adopted for commercial purposes, the company's FPGAs and tools also enjoy

use by universities to support FPGA-based computing research.

Although engineered for varied applications, one of the distinct capabilities of Altera’s

Generation 10 FPGAs is its support for high-level programming languages. While

traditional fixed point FPGAs support only Verilog, VHDL, and Simulink, incorporation of

the floating point functionalities has enabled Altera’s FPGAs to be more compatible with

higher level programming languages, such as C, OpenCL, MATLAB, and Simulink. With

advanced functionalities and application diversity brought on by hard floating point in its

FPGAs, Altera’s recent innovation has opened up new avenues of growth for the company.

Customer Acquisition

Intensive internal research and development, a strong commitment to innovation, and the

ability to deliver high-quality products meeting customer requirements are the three

pillars of Altera’s customer acquisition strategy. Apart from gauging customer

requirements for FPGAs for over a decade and regularly upgrading its products with

enhanced capabilities, the company has continued to identify growth opportunities.

Altera’s FPGAs were used by Microsoft Research to run convolutional neural network

(CNN) algorithms, and floating point algorithms and by Baidu (in September 2014) to

generate faster, more accurate search results. These are just two examples of the

company’s visionary innovation and understanding of customer requirements. Altera’s

capability to provide unparalleled processing speed in an FPGA that offers a new level of

computational productivity, without the need for increased power, has helped the

company build a robust third-party board supplier base, which providers such as Bittware,



Gidel, Nallatech, and ReFLEX CES. These organizations highly appreciate Altera’s Arria 10

FPGAs for the company’s undisputed capability in DSP and high-performance computing

(HPC). With the growing need for hard floating point, data center acceleration in HPC

applications, Altera’s Stratix 10 range is also likely to be widely adopted, thereby adding

to the company’s growth potential going forward.

Technology Licensing

Altera expects use of hard floating point FPGAs to become the industry standard in the

coming years. Frost & Sullivan also notes that Altera's other competitors are already

striving to incorporate floating point operators into their FPGAs. Although Altera is aware

of the increasing demand for its floating point FPGAs, the company has designed a flexible

licensing policy. The floating point architecture is fully available through Altera’s Quartus

II design suite, and also available through more specialized tools such as the Altera SDK

for OpenCL, and DSPBuilder design tools.The strategy not only offers Altera’s customers

flexibility of use, but also enhances the company’s brand visibility.

Conclusion

Altera’s work at engineering IEEE-754 floating-point on its Arria 10 and Stratix 10 FPGAs

thereby adding advanced design capabilities, empowers customers with the FPGA’s

unrivalled performance. Such holistic solutions, along with the innovative Hyperflex

architecture, not only ensures enhanced performance, but also supports varied

applications in next-generation wireless systems, medical imaging, radar, and data

centers. With a flexible technology licensing policy that allows customers to buy FPGAs

and license the supporting tools without having to pay extra royalty or per unit cost,

Altera has a well-articulated strategy for customer acquisition. With its strong overall

performance, Altera Corporation has earned the 2015 Frost & Sullivan Global Technology

Innovation Leadership Award.



Significance of Technology Innovation Ultimately, growth in any organization depends upon finding new ways to excite the

market, and upon maintaining a long-term commitment to innovation. At its core,

technology innovation or any other type of innovation can only be sustained with

leadership in three key areas: understanding demand, nurturing the brand, and

differentiating from the competition.

Understanding Technology Innovation

Technology innovation begins with a spark of creativity that is systematically pursued,

developed, and commercialized. This spark can result from a successful partnership, a

productive in-house innovation group, or the mind of a single individual. Regardless of the

source, the success of any new technology is ultimately determined by its innovativeness

and its impact on the business as a whole.



Key Benchmarking Criteria

For the Global Technology Innovation Leadership Award, Frost & Sullivan analysts

independently evaluated two key factors—Technology Attributes and Future Business

Value—according to the criteria identified below.

Technology Attributes

Criterion 1: Industry Impact

Criterion 2: Product Impact

Criterion 3: Scalability

Criterion 4: Visionary Innovation

Criterion 5: Application Diversity

Future Business Value

Criterion 1: Financial Performance

Criterion 2: Customer Acquisition

Criterion 3: Technology Licensing

Criterion 4: Brand Loyalty

Criterion 5: Human Capital

Best Practice Award Analysis for Altera Corporation

Decision Support Scorecard

To support its evaluation of best practices across multiple business performance

categories, Frost & Sullivan employs a customized Decision Support Scorecard. This tool

allows our research and consulting teams to objectively analyze performance, according to

the key benchmarking criteria listed in the previous section, and to assign ratings on that

basis. The tool follows a 10-point scale that allows for nuances in performance evaluation;

ratings guidelines are illustrated below.

RATINGS GUIDELINES

The Decision Support Scorecard is organized by Technology Attributes and Future

Business Value (i.e., the overarching categories for all 10 benchmarking criteria; the

definitions for each criteria are provided beneath the scorecard). The research team

confirms the veracity of this weighted scorecard through sensitivity analysis, which

confirms that small changes to the ratings for a specific criterion do not lead to a

significant change in the overall relative rankings of the companies.



The results of this analysis are shown below. To remain unbiased and to protect the

interests of all organizations reviewed, we have chosen to refer to the other key players

as Competitor 2 and Competitor 3.

DECISION SUPPORT SCORECARD FOR TECHNOLOGY INNOVATION AWARD

Measurement of 1–10 (1 = poor; 10 = excellent)

Technology Innovation

Technology

Attributes

Future

Business Value Average Rating

Altera Corporation 9.6 9.6 9.6

Competitor 2 8.5 8.3 8.4

Competitor 3 7.6 7.6 7.6


Criterion 1: Industry Impact

Requirement: Technology enables the pursuit of groundbreaking new ideas, contributing

to the betterment of the entire industry

Criterion 2: Product Impact

Requirement: Specific technology helps enhance features and functionality of the entire

product line for the company

Criterion 3: Scalability

Requirement: Technology is scalable, enabling new generations of products over time,

with increasing levels of quality and functionality

Criterion 4: Visionary Innovation

Requirement: Specific new technology represents true innovation based on a deep

understanding of future needs and applications

Criterion 5: Application Diversity

Requirement: New technology serves multiple products, multiple applications, and

multiple user environments


Criterion 1: Financial Performance

Requirement: High potential for strong financial performance in terms of revenues,

operating margins and other relevant financial metrics

Criterion 2: Customer Acquisition

Requirement: Specific technology enables acquisition of new customers, even as it

enhances value to current customers



Criterion 3: Technology Licensing

Requirement: New technology displays great potential to be licensed across many sectors

and applications, thereby driving incremental revenue streams

Criterion 4: Brand Loyalty

Requirement: New technology enhances the company’s brand, creating and/or nurturing

brand loyalty

Criterion 5: Human Capital

Requirement: Customer impact is enhanced through the leverage of specific technology,

translating into positive impact on employee morale and retention

Decision Support Matrix

Once all companies have been evaluated according to the Decision Support Scorecard,

analysts can then position the candidates on the matrix shown below, enabling them to

visualize which companies are truly breakthrough and which ones are not yet operating at

best-in-class levels.

DECISION SUPPORT MATRIX FOR TECHNOLOGY INNOVATION AWARD

High

Low

Low High



Altera

Competitor 2

Competitor 3



The Intersection between 360-Degree Research and Best

Practices Awards

Research Methodology

Frost & Sullivan’s 360-degree research

methodology represents the analytical

rigor of our research process. It offers a

360-degree-view of industry challenges,

trends, and issues by integrating all 7 of

Frost & Sullivan's research methodologies.

Too often, companies make important

growth decisions based on a narrow

understanding of their environment,

leading to errors of both omission and

commission. Successful growth strategies

are founded on a thorough understanding

of market, technical, economic, financial,

customer, best practices, and demographic

analyses. The integration of these research

disciplines into the 360-degree research

methodology provides an evaluation

platform for benchmarking industry players and for identifying those performing at best-

in-class levels.

360-DEGREE RESEARCH: SEEING ORDER IN

THE CHAOS

Technology

Obsolescence

Disruptive

Technologies

New

Applications

CEO

Demographics

Needs

and

PerceptionsSegmentation

Buying

Behavior

Branding

and

Positioning

Competitive

Benchmarking

Emerging

Competition

Competitive

Strategy

Capital

Investments

Availability

of

Capital

Country

Risk

Economic

Trends

Crowd

Sourcing

Growth

Strategies

Career

Development

Growth

Implementation

Industry

Evolution

New Vertical

Markets

Industry

Expansion

Industry

Convergence

Emerging

Technologies

Smart Cities

Sustainability

New Business

Cultures

GeoPolitical

Stability



Best Practices Recognition: 10 Steps to Researching,

Identifying, and Recognizing Best Practices

Frost & Sullivan Awards follow a 10-step process to evaluate Award candidates and assess

their fit with select best practice criteria. The reputation and integrity of the Awards are

based on close adherence to this process.

STEP OBJECTIVE KEY ACTIVITIES OUTPUT

1 Monitor, target, and screen

Identify Award recipient candidates from around the globe

• Conduct in-depth industry research

• Identify emerging sectors • Scan multiple geographies

Pipeline of candidates who potentially meet all best-practice criteria

2 Perform 360-degree research

Perform comprehensive, 360-degree research on all candidates in the pipeline

• Interview thought leaders and industry practitioners

• Assess candidates’ fit with best-practice criteria

• Rank all candidates

Matrix positioning all candidates’ performance relative to one another

3

Invite thought leadership in best practices

Perform in-depth examination of all candidates

• Confirm best-practice criteria • Examine eligibility of all candidates

• Identify any information gaps

Detailed profiles of all ranked candidates

4

Initiate research director review

Conduct an unbiased evaluation of all candidate profiles

• Brainstorm ranking options • Invite multiple perspectives on candidates’ performance

• Update candidate profiles

Final prioritization of all eligible candidates and companion best-practice positioning paper

5

Assemble panel of industry experts

Present findings to an expert panel of industry thought leaders

• Share findings • Strengthen cases for candidate eligibility

• Prioritize candidates

Refined list of prioritized Award candidates

6

Conduct global industry review

Build consensus on Award candidates’ eligibility

• Hold global team meeting to review all candidates

• Pressure-test fit with criteria • Confirm inclusion of all eligible candidates

Final list of eligible Award candidates, representing success stories worldwide

7 Perform quality check

Develop official Award consideration materials

• Perform final performance benchmarking activities

• Write nominations • Perform quality review

High-quality, accurate, and creative presentation of nominees’ successes

8

Reconnect with panel of industry experts

Finalize the selection of the best-practice Award recipient

• Review analysis with panel • Build consensus • Select winner

Decision on which company performs best against all best-practice criteria

9 Communicate recognition

Inform Award recipient of Award recognition

• Present Award to the CEO • Inspire the organization for continued success

• Celebrate the recipient’s performance

Announcement of Award and plan for how recipient can use the Award to enhance the brand

10 Take strategic action

Upon licensing, company may share Award news with stakeholders and customers

• Coordinate media outreach • Design a marketing plan • Assess Award’s role in future strategic planning

Widespread awareness of recipient’s Award status among investors, media personnel, and employees



About Frost & Sullivan

Frost & Sullivan, the Growth Partnership Company, enables clients to accelerate growth

and achieve best in class positions in growth, innovation and leadership. The company's

Growth Partnership Service provides the CEO and the CEO's Growth Team with disciplined

research and best practice models to drive the generation, evaluation and implementation

of powerful growth strategies. Frost & Sullivan leverages almost 50 years of experience in

partnering with Global 1000 companies, emerging businesses and the investment

community from 31 offices on six continents. To join our Growth Partnership, please visit

http://www.frost.com.

2015 global field programmable gate array (fpga ... · gate array (fpga) product lines; these are...

Documents