Eduardo Cuervo - DukeAruna Balasubramanian - U Mass AmherstDae-ki Cho - UCLAAlec Wolman, Stefan Saroiu, Ranveer Chandra, Paramvir Bahl – Microsoft Research

CPU performance during same period: 246X

A solution to the battery problem seems unlikely

Just 2X in 15 years

Augmented Reality

Speech Recognition and Synthesis

Interactive Games

Slow, Limited or Inaccurate

Too CPU intensive Limit

ed

Power Intensive

Not on par with desktop

counterparts

Remote execution can reduce energy consumption

Challenges: What should be offloaded? How to dynamically decide when to offload? How to minimize the required programmer

effort?

MAUI Contributions: Combine extensive profiling with an ILP

solver Makes dynamic offload decisions Optimize for energy reduction Profile: device, network, application

Leverage modern language runtime (.NET CLR) To simplify program partitioning Reflection, serialization, strong typing

MotivationMAUI system design

MAUI proxy MAUI profiler MAUI solver

EvaluationConclusion

Maui serverSmartphone

Application

Client Proxy

Profiler

Solver

Maui Runtime

Server Proxy

Profiler

Solver

Maui Runtime

Application

RPC

RPC

Maui Controller

Goal: make it dead-simple to MAUI-ify apps Build app as a standalone phone app Add .NET attributes to indicate

“remoteable” Follow a simple set of rules

Portability: Mobile (ARM) vs Server (x86) .NET Framework Common Intermediate

Language Type-Safety and Serialization:

Automate state extraction Reflection:

Identifies methods with [Remoteable] tag Automates generation of RPC stubs

Maui serverSmartphone

Application

Client Proxy

Profiler

Solver

Maui Runtime

Server Proxy

Profiler

Solver

Maui Runtime

Application

RPC

RPC

Maui Controller

Intercepts Application CallsSynchronizes State

Chooses local or remote

Handles Errors

Provides runtime information

ProfilerProfilerCallgraphCallgraph

Execution TimeExecution Time

State sizeState size

Network LatencyNetwork Latency

Network BandwidthNetwork Bandwidth

Device ProfileDevice ProfileCPU CyclesCPU Cycles

Network Power CostNetwork DelayComputational Delay

Computational Power Cost

Computational Delay

An

nota

ted

C

allg

rap

h

B900 mJ15ms

C5000 mJ3000 ms

1000mJ25000 m

J D15000 mJ12000 ms

1000

0 m

J

A

Computation energy and delay for execution

Energy and delay for state transfer

A sample callgraph

FindMatch

900 mJ

InitializeFaceRecognizer

5000 mJ

1000mJ25000 m

J DetectAndExtract Faces15000 mJ

1000

0 m

J

UserInterface

Yes! – This simple example from Face Recognition app shows why local analysis fails.

Cheaper to do local

FindMatch

900 mJ

InitializeFaceRecognizer

5000 mJ

1000mJ25000 m

J DetectAndExtract Faces15000 mJ

1000

0 m

J

UserInterface

Yes! – This simple example from Face Recognition app shows why local analysis fails.

Cheaper to do local

Cheaper to do local

FindMatch

InitializeFace

Recognizer

1000mJ

DetectAndExtract Faces

UserInterface 25900mJ

Cheaper to offload

Adapt to: Network Bandwidth/Latency Changes Variability on method’s computational

requirements

Experiment: Modified off the shelf arcade game

application Physics Modeling (homing missiles) Evaluated under different latency settings

DoLevel

HandleMissilesHandleMissiles

DoFrame

HandleEnemies

HandleBonuses

11KB + missiles

11KB

+

mis

sile

s

missiles

*Missiles take around 60 bytes each

11KB + missiles

Required state is smaller

Complexity increases with # of missiles

DoLevel

HandleMissiles

DoFrame

HandleEnemies

HandleBonuses

*Missiles take around 60 bytes each

Zero MissilesLow latency (RTT < 10ms)

Computation cost is close to zero

Offload starting at DoLevel

DoLevel

HandleMissiles

DoFrame

HandleEnemies

HandleBonuses

*Missiles take around 60 bytes each

5 MissilesSome latency (RTT = 50ms)

Most of the computation cost

Very expensive to offload everything

Little state to offload

Only offload Handle Missiles

MotivationMAUI system design

MAUI proxy MAUI profiler MAUI solver

EvaluationConclusion

Platform Windows Mobile 6.5 .NET Framework 3.5 HTC Fuze Smartphone Monsoon power monitor

Applications Chess Face Recognition Arcade Game Voice-based translator

How much can MAUI reduce energy consumption?

How much can MAUI improve performance?

Can MAUI Run Resource-Intensive Applications?

Big savings even on 3G

An order of magnitude

improvement on Wi-Fi

Face Recognizer

Improvement of around an order of magnitude

Face Recognizer

Up to 40% energy savings on Wi-Fi

Solver would decide not to offloadArcade Game

CPU Intensive even on a Core 2 Duo PC

Can be run on the phone with MAUI

Translator

MotivationMAUI system design

MAUI proxy MAUI profiler MAUI solver

EvaluationConclusion

MAUI enables developers to: Bypass the resource limitations of handheld devices Low barrier entry: simple program annotations

For a resource-intensive application MAUI reduced energy consumed by an order of

magnitude MAUI improved application performance similarly

MAUI adapts to: Changing network conditions Changing applications CPU demands

http://research.microsoft.com/en-us/projects/maui/

ecuervo@cs.duke.edu