Share
Embed
GreenWeb: Language Extensions for Energy-Efficient Mobile Web Computing
PLDI 2016 Presentation
Yuhao Zhu
June 15, 2016
Transcript
131 slides
-
GreenWeb: Language Extensions for Energy-Efficient Mobile Web Computing Yuhao Zhu
The University of Texas at Austin with Vijay Janapa Reddi PLDI 2016 1 -
0 30 60 90 120 2011 2012 2013 2014 2015
2016 2 Source: BIA/Kelsey Search Volume (B) Web: Mobile Overtaking Desktop -
0 30 60 90 120 2011 2012 2013 2014 2015
2016 2 Source: BIA/Kelsey Search Volume (B) Mobile Desktop Web: Mobile Overtaking Desktop -
0 30 60 90 120 2011 2012 2013 2014 2015
2016 2 Source: BIA/Kelsey Search Volume (B) Mobile Desktop Web: Mobile Overtaking Desktop -
Energy Concern Among Mobile Developers 4 [ICSE 2016] Manotas et
al., “An Empirical Study of Practitioners’ Perspectives on Green Software Engineering” -
Energy Concern Among Mobile Developers 4 Percentage (%) 0 25
50 75 100 Mobile Desktop Data Center Never/Rarely Sometimes Often/Almost Always “My applications have requirements about energy usage.” [ICSE 2016] Manotas et al., “An Empirical Study of Practitioners’ Perspectives on Green Software Engineering” -
Energy Concern Among Mobile Developers 4 Percentage (%) 0 25
50 75 100 Mobile Desktop Data Center Never/Rarely Sometimes Often/Almost Always “My applications have requirements about energy usage.” [ICSE 2016] Manotas et al., “An Empirical Study of Practitioners’ Perspectives on Green Software Engineering” -
Energy Concern Among Mobile Developers 4 Percentage (%) 0 25
50 75 100 Mobile Desktop Data Center Never/Rarely Sometimes Often/Almost Always “My applications have requirements about energy usage.” [ICSE 2016] Manotas et al., “An Empirical Study of Practitioners’ Perspectives on Green Software Engineering” -
Developers are Willing to Make Trade-offs 5 [ICSE 2016] Manotas
et al., “An Empirical Study of Practitioners’ Perspectives on Green Software Engineering” -
Developers are Willing to Make Trade-offs 5 Percentage (%) 0
25 50 75 100 Mobile Never/Rarely Sometimes Often/Almost Always “I'm willing to sacrifice performance, etc. for reduced energy usage.” [ICSE 2016] Manotas et al., “An Empirical Study of Practitioners’ Perspectives on Green Software Engineering” -
Developers are Willing to Make Trade-offs 5 Percentage (%) 0
25 50 75 100 Mobile Never/Rarely Sometimes Often/Almost Always “I'm willing to sacrifice performance, etc. for reduced energy usage.” [ICSE 2016] Manotas et al., “An Empirical Study of Practitioners’ Perspectives on Green Software Engineering” -
Quality-of-service Energy-efficiency Conflicting requirements 7 GreenWeb Programming language support for
balancing energy-efficiency and QoS in mobile Web computing -
8 GreenWeb Programming language support for balancing energy-efficiency and QoS
in mobile Web computing -
GreenWeb 8 GreenWeb Programming language support for balancing energy-efficiency and
QoS in mobile Web computing -
9 ▸ Runtime that saves energy while meeting the QoS
constraints GreenWeb: Language for Energy-Efficiency ▸ Language abstractions for expressing QoS -
9 ▸ Runtime that saves energy while meeting the QoS
constraints ▸ Result in 60% energy savings on real hardware/software implementations GreenWeb: Language for Energy-Efficiency ▸ Language abstractions for expressing QoS -
9 ▸ Runtime the QoS constraints ▸ Result hardware/software implementations
GreenWeb: Language for Energy-Efficiency ▸ Language abstractions for expressing QoS -
11 Performance QoS Experience [OSDI 1996] Y. Endo et al.,
“Using Latency to Evaluate Interactive System Performance.” Understanding Mobile Web QoS -
11 Performance QoS Experience [OSDI 1996] Y. Endo et al.,
“Using Latency to Evaluate Interactive System Performance.” Understanding Mobile Web QoS Too slow -
11 Performance QoS Experience Unusable [OSDI 1996] Y. Endo et
al., “Using Latency to Evaluate Interactive System Performance.” Understanding Mobile Web QoS Too slow -
11 Performance QoS Experience Unusable Tolerable [OSDI 1996] Y. Endo
et al., “Using Latency to Evaluate Interactive System Performance.” Understanding Mobile Web QoS Too slow -
11 Performance QoS Experience Unusable Tolerable [OSDI 1996] Y. Endo
et al., “Using Latency to Evaluate Interactive System Performance.” Understanding Mobile Web QoS Too slow Diminishing Returns -
11 Performance QoS Experience Unusable Tolerable Imperceptible [OSDI 1996] Y.
Endo et al., “Using Latency to Evaluate Interactive System Performance.” Understanding Mobile Web QoS Too slow Diminishing Returns -
12 Performance QoS Experience Unusable Tolerable Imperceptible Understanding Mobile Web
QoS Energy [OSDI 1996] Y. Endo et al., “Using Latency to Evaluate Interactive System Performance.” -
12 Performance QoS Experience Unusable Tolerable Imperceptible Understanding Mobile Web
QoS Energy [OSDI 1996] Y. Endo et al., “Using Latency to Evaluate Interactive System Performance.” -
12 Performance QoS Experience Unusable Tolerable Imperceptible Understanding Mobile Web
QoS Energy [OSDI 1996] Y. Endo et al., “Using Latency to Evaluate Interactive System Performance.” -
12 Performance QoS Experience Unusable Tolerable Imperceptible Understanding Mobile Web
QoS Energy [OSDI 1996] Y. Endo et al., “Using Latency to Evaluate Interactive System Performance.” -
12 Performance QoS Experience Unusable Tolerable Imperceptible Understanding Mobile Web
QoS “Negative” Energy consumption Energy [OSDI 1996] Y. Endo et al., “Using Latency to Evaluate Interactive System Performance.” -
12 Performance QoS Experience Unusable Tolerable Imperceptible Understanding Mobile Web
QoS Energy [OSDI 1996] Y. Endo et al., “Using Latency to Evaluate Interactive System Performance.” -
12 Performance QoS Experience Unusable Tolerable Imperceptible Understanding Mobile Web
QoS Energy [OSDI 1996] Y. Endo et al., “Using Latency to Evaluate Interactive System Performance.” -
13 Performance QoS Experience Unusable Tolerable Imperceptible Abstracting Mobile Web
QoS ▸ Performance metric ▹ Frame latency vs. Frame throughput -
13 Performance QoS Experience Unusable Tolerable Imperceptible Abstracting Mobile Web
QoS ▸ Performance metric ▹ Frame latency vs. Frame throughput QoS Type -
13 Performance QoS Experience Unusable Tolerable Imperceptible Abstracting Mobile Web
QoS ▸ Performance metric ▹ Frame latency vs. Frame throughput ▸ Threshold performance values ▹ Imperceptible target vs. Usable target QoS Type -
13 Performance QoS Experience Unusable Tolerable Imperceptible Abstracting Mobile Web
QoS ▸ Performance metric ▹ Frame latency vs. Frame throughput ▸ Threshold performance values ▹ Imperceptible target vs. Usable target QoS Type QoS Target -
<html> <head> <script> function animateMove() { /* Animation code omitted
*/ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> 14 Expressing Mobile Web QoS -
<html> <head> <script> function animateMove() { /* Animation code omitted
*/ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> 14 Expressing Mobile Web QoS element -
<html> <head> <script> function animateMove() { /* Animation code omitted
*/ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> 14 Expressing Mobile Web QoS element event -
<html> <head> <script> function animateMove() { /* Animation code omitted
*/ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> 14 Expressing Mobile Web QoS element event -
<html> <head> <script> function animateMove() { /* Animation code omitted
*/ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> 14 Expressing Mobile Web QoS Expressing QoS at an event granularity element event -
<script> function animateMove() { /* Animation code omitted */ }
</script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> <style> </style> <html> <head> 15 Expressing Mobile Web QoS Annotation element event -
<script> function animateMove() { /* Animation code omitted */ }
</script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> <style> </style> <html> <head> 15 Expressing Mobile Web QoS Annotation element event div { } ontouchend -
{ : Type, Target} <script> function animateMove() { /* Animation
code omitted */ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> <style> </style> <html> <head> 15 Expressing Mobile Web QoS Annotation element event div { } ontouchend: throughput, low; -
{ : Type, Target} <script> function animateMove() { /* Animation
code omitted */ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> <style> </style> <html> <head> 15 Expressing Mobile Web QoS Annotation element event div { } ontouchend: throughput, low; -
{ : Type, Target} <script> function animateMove() { /* Animation
code omitted */ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> <style> </style> <html> <head> 15 Expressing Mobile Web QoS Annotation element event div { } ontouchend: throughput, low; -
{ : Type, Target} <script> function animateMove() { /* Animation
code omitted */ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> <style> </style> <html> <head> 15 Expressing Mobile Web QoS Annotation function newAnimateMove() { /* New animation code */ } element event div { } ontouchend: throughput, low; -
{ : Type, Target} <script> function animateMove() { /* Animation
code omitted */ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> <style> </style> <html> <head> 15 Expressing Mobile Web QoS Annotation function newAnimateMove() { /* New animation code */ } Implementation independent element event div { } ontouchend: throughput, low; -
{ : Type, Target} <script> function animateMove() { /* Animation
code omitted */ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> <style> </style> <html> <head> 15 Expressing Mobile Web QoS Annotation function newAnimateMove() { /* New animation code */ } Implementation independent element event div { } ontouchend: throughput, low; -
{ : Type, Target} <script> function animateMove() { /* Animation
code omitted */ } </script> </head> <body> <div ontouchend=“animateMove()”> <div/> <!— other elements --> </body> </html> <style> </style> <html> <head> 15 Expressing Mobile Web QoS Annotation function newAnimateMove() { /* New animation code */ } Implementation independent Non-interfering w.r.t. functionality element event div { } ontouchend: throughput, low; -
16 Original application GreenWeb- annotated application GreenWeb Annotation Process Automatic
Annotation? ▸ AutoGreen: automatically reasons about and inserts GreenWeb annotations -
16 GreenWeb- annotated application GreenWeb Annotation Process Automatic Annotation? ▸
AutoGreen: automatically reasons about and inserts GreenWeb annotations DOM Tree -
▸ AutoGreen: automatically reasons about and inserts GreenWeb annotations 17
GreenWeb- annotated application GreenWeb Annotation Process -
▸ AutoGreen: automatically reasons about and inserts GreenWeb annotations 17
GreenWeb- annotated application GreenWeb Annotation Process Callback Instrumentation -
▸ AutoGreen: automatically reasons about and inserts GreenWeb annotations 17
GreenWeb- annotated application GreenWeb Annotation Process QoS Information Event Profiling Callback Instrumentation -
▸ AutoGreen: automatically reasons about and inserts GreenWeb annotations 17
GreenWeb- annotated application GreenWeb Annotation Process QoS Information Event Profiling Annotation Generation Callback Instrumentation -
▸ Language abstractions for expressing QoS 18 ▸ Runtime the
QoS constraints ▸ Result hardware/software implementations GreenWeb: Language for Energy-Efficiency -
▸ Language abstractions 18 ▸ Runtime that saves energy while
meeting the QoS constraints ▸ Result hardware/software implementations GreenWeb: Language for Energy-Efficiency -
19 GreenWeb Runtime Overview Frame Event Enforcing event-level QoS at
the frame-level energy-efficiently Runtime Objective QoS Annotations -
QoS type: latency QoS target: 16 ms 19 GreenWeb Runtime
Overview Frame Event Enforcing event-level QoS at the frame-level energy-efficiently Runtime Objective -
QoS type: latency QoS target: 16 ms 19 GreenWeb Runtime
Overview Frame Event Enforcing event-level QoS at the frame-level energy-efficiently Runtime Objective -
QoS type: latency QoS target: 16 ms 19 GreenWeb Runtime
Overview Frame Event Enforcing event-level QoS at the frame-level energy-efficiently Runtime Objective -
QoS type: latency QoS target: 16 ms 19 GreenWeb Runtime
Overview Frame Event 16 ms Enforcing event-level QoS at the frame-level energy-efficiently Runtime Objective -
QoS type: latency QoS target: 16 ms throughput 19 GreenWeb
Runtime Overview Frame Event 16 ms Enforcing event-level QoS at the frame-level energy-efficiently Runtime Objective -
QoS type: latency QoS target: 16 ms throughput 19 GreenWeb
Runtime Overview Frame Event Frame Frame 16 ms 16 ms 16 ms Enforcing event-level QoS at the frame-level energy-efficiently Runtime Objective -
19 GreenWeb Runtime Overview Frame Event Frame Frame Time Event
16 ms 16 ms 16 ms Enforcing event-level QoS at the frame-level energy-efficiently Runtime Objective QoS target: 2 s -
19 GreenWeb Runtime Overview Frame Event Frame Frame Time Event
Frame 16 ms 16 ms 16 ms Enforcing event-level QoS at the frame-level energy-efficiently Runtime Objective 2 s QoS target: 2 s -
19 GreenWeb Runtime Overview Frame Event Frame Frame Time Event
Frame 16 ms 16 ms 16 ms Enforcing event-level QoS at the frame-level energy-efficiently { Runtime Objective 2 s QoS target: 2 s -
19 GreenWeb Runtime Overview Frame Event Frame Frame Time Event
Frame 16 ms 16 ms 16 ms Enforcing event-level QoS at the frame-level energy-efficiently Frame Association { Runtime Objective 1 2 s -
19 GreenWeb Runtime Overview Frame Event Frame Frame Time Event
Frame 16 ms 16 ms 16 ms Enforcing event-level QoS at the frame-level energy-efficiently Frame Association Frame Scheduling { Runtime Objective 1 2 2 s -
21 Frame Association S S L P C Frame Browser
Process Renderer Process GPU Process Event -
21 Frame Association S S L P C Frame Browser
Process Renderer Process GPU Process Event Main Thread Compositor Thread -
21 Frame Association S S L P C Frame Browser
Process Renderer Process GPU Process Event Main Thread Compositor Thread IPC Inter-thread Message -
22 Frame Association S S L P C Frame Browser
Process Renderer Process GPU Process Event Frame S S L P C -
22 Frame Association S S L P C Frame Browser
Process Renderer Process GPU Process Event Frame S S L P C Distribute QoS information along with the communication messages -
Choices of Energy-saving Techniques 23 GreenWeb can support a range
of energy saving techniques -
Choices of Energy-saving Techniques 23 GreenWeb can support a range
of energy saving techniques ▹Dynamic resolution scaling [MobiCom 2015] ▹Power-saving display colors [MobiSys 2012] ▹Selective resource loading [NSDI 2015] -
Choices of Energy-saving Techniques 23 GreenWeb can support a range
of energy saving techniques ▹Dynamic resolution scaling [MobiCom 2015] ▹Power-saving display colors [MobiSys 2012] ▹Selective resource loading [NSDI 2015] ▹ACMP-based hardware mechanism -
Energy Consumption Performance Big Core Small Core ACMP-based Hardware Substrate
24 ▸ Asymmetric Chip-multiprocessor, a.k.a., Big/Little architecture -
Energy Consumption Performance Big Core Small Core ACMP-based Hardware Substrate
24 ▸ Asymmetric Chip-multiprocessor, a.k.a., Big/Little architecture Frequency Levels -
Energy Consumption Performance Big Core Small Core ACMP-based Hardware Substrate
24 ▸ Asymmetric Chip-multiprocessor, a.k.a., Big/Little architecture ▸ Already used in commodity devices (e.g., Samsung Galaxy S6) Frequency Levels -
Energy Consumption Performance Big Core Small Core ACMP-based GreenWeb Runtime
25 ▸ Provide just enough energy to meet QoS constraints -
Energy Consumption Performance Big Core Small Core ACMP-based GreenWeb Runtime
25 ▸ Provide just enough energy to meet QoS constraints div {ontouchend: latency, 16 ms} -
Energy Consumption Performance Big Core Small Core ACMP-based GreenWeb Runtime
25 ▸ Provide just enough energy to meet QoS constraints 16 ms div {ontouchend: latency, 16 ms} -
Energy Consumption Performance Big Core Small Core ACMP-based GreenWeb Runtime
25 ▸ Provide just enough energy to meet QoS constraints 16 ms div {ontouchend: latency, 16 ms} -
Energy Consumption Performance Big Core Small Core ACMP-based GreenWeb Runtime
25 ▸ Provide just enough energy to meet QoS constraints ▸ Event-based scheduling [HPCA 2015] -
Energy Consumption Performance ACMP-based GreenWeb Runtime 26 ▸ Provide just
enough energy to meet QoS constraints ▸ Event-based scheduling [HPCA 2015] Execution Time = [PLDI 2003] Xie, et al., “Compile-Time Dynamic Voltage Scaling Settings: Opportunities and Limits” Energy = -
Energy Consumption Performance ACMP-based GreenWeb Runtime 26 ▸ Provide just
enough energy to meet QoS constraints ▸ Event-based scheduling [HPCA 2015] Execution Time = Tmemory + [PLDI 2003] Xie, et al., “Compile-Time Dynamic Voltage Scaling Settings: Opportunities and Limits” Energy = -
Tcpu Energy Consumption Performance ACMP-based GreenWeb Runtime 26 ▸ Provide
just enough energy to meet QoS constraints ▸ Event-based scheduling [HPCA 2015] Execution Time = Tmemory + [PLDI 2003] Xie, et al., “Compile-Time Dynamic Voltage Scaling Settings: Opportunities and Limits” Energy = -
Energy Consumption Performance ACMP-based GreenWeb Runtime 26 ▸ Provide just
enough energy to meet QoS constraints ▸ Event-based scheduling [HPCA 2015] Execution Time = Tmemory + Ncycles / f [PLDI 2003] Xie, et al., “Compile-Time Dynamic Voltage Scaling Settings: Opportunities and Limits” Energy = -
Energy Consumption Performance ACMP-based GreenWeb Runtime 26 ▸ Provide just
enough energy to meet QoS constraints ▸ Event-based scheduling [HPCA 2015] Execution Time = Tmemory + Ncycles / f [PLDI 2003] Xie, et al., “Compile-Time Dynamic Voltage Scaling Settings: Opportunities and Limits” Energy = -
Energy Consumption Performance ACMP-based GreenWeb Runtime 26 ▸ Provide just
enough energy to meet QoS constraints ▸ Event-based scheduling [HPCA 2015] Execution Time = Tmemory + Ncycles / f [PLDI 2003] Xie, et al., “Compile-Time Dynamic Voltage Scaling Settings: Opportunities and Limits” Energy = Execution Time x Power -
▸ Language abstractions 27 ▸ Runtime that saves energy while
meeting the QoS constraints ▸ Result hardware/software implementations GreenWeb: Language for Energy-Efficiency -
▸ Language abstractions 27 ▸ Runtime the QoS constraints ▸
Result in 60% energy savings on real hardware/software implementations GreenWeb: Language for Energy-Efficiency -
Real Hardware/Software Setup 28 ODroid XU+E development board, which contains
an Exynos 5410 SoC used in Samsung Galaxy S4. -
Real Hardware/Software Setup 28 ODroid XU+E development board, which contains
an Exynos 5410 SoC used in Samsung Galaxy S4. Implementation incorporated into Chrome running on Android. -
Real Hardware/Software Setup 28 ODroid XU+E development board, which contains
an Exynos 5410 SoC used in Samsung Galaxy S4. Implementation incorporated into Chrome running on Android. UI-level record and replay for reproducibility. [ISPASS’15] -
Evaluation ▸Baseline Mechanisms ▹Highest performance (Perf) — Standard to guarantee
responsiveness ▹Interactive governor (Interactive) — Android default 29 29 -
Evaluation ▸Baseline Mechanisms ▹Highest performance (Perf) — Standard to guarantee
responsiveness ▹Interactive governor (Interactive) — Android default 29 ▸Metrics ▹Energy Saving ▹QoS Violation 29 -
Evaluation ▸Baseline Mechanisms ▹Highest performance (Perf) — Standard to guarantee
responsiveness ▹Interactive governor (Interactive) — Android default 29 ▸Metrics ▹Energy Saving ▹QoS Violation 29 ▸Applications ▹Top webpages (e.g., www.amazon.com) ▹Web Apps based on popular frameworks (e.g., Todo List) -
30 Norm. Energy 0.0 0.3 0.5 0.8 1.0 CamanJS Craigslist
Paperjs Goo Google Todo CNet BBC LZMA-JS Amazon W3School MSN GreenWeb Interactive Perf Evaluation Results -
31 Norm. Energy 0.0 0.3 0.5 0.8 1.0 CamanJS Craigslist
Paperjs Goo Google Todo CNet BBC LZMA-JS Amazon W3School MSN GreenWeb Interactive Perf Evaluation Results -
32 Norm. Energy 0.0 0.3 0.5 0.8 1.0 CamanJS Craigslist
Paperjs Goo Google Todo CNet BBC LZMA-JS Amazon W3School MSN GreenWeb Interactive Perf Evaluation Results -
33 Evaluation Results QoS Violations (%) 0.0 0.8 1.5 2.3
3.0 CamanJS Craigslist Paperjs Goo Google Todo CNet BBC LZMA-JS Amazon W3School MSN Norm. Energy 0.0 0.3 0.5 0.8 1.0 CamanJS Craigslist Paperjs Goo Google Todo CNet BBC LZMA-JS Amazon W3School MSN GreenWeb Interactive Perf -
Norm. Energy 0.0 0.3 0.5 0.8 1.0 CamanJS Craigslist Paperjs
Goo Google Todo CNet BBC LZMA-JS Amazon W3School MSN GreenWeb Interactive Perf 34 Evaluation Results QoS Violations (%) 0.0 0.8 1.5 2.3 3.0 CamanJS Craigslist Paperjs Goo Google Todo CNet BBC LZMA-JS Amazon W3School MSN No QoS Violations -
Norm. Energy 0.0 0.3 0.5 0.8 1.0 CamanJS Craigslist Paperjs
Goo Google Todo CNet BBC LZMA-JS Amazon W3School MSN GreenWeb Interactive Perf 34 Evaluation Results QoS Violations (%) 0.0 0.8 1.5 2.3 3.0 CamanJS Craigslist Paperjs Goo Google Todo CNet BBC LZMA-JS Amazon W3School MSN 29.2% - 66.0% energy savings, 0.8% more QoS violations No QoS Violations -
35 GreenWeb Programming language support for balancing energy-efficiency and QoS
in mobile Web computing -
35 GreenWeb Programming language support for balancing energy-efficiency and QoS
in mobile Web computing Abstraction Express QoS constraints -
35 GreenWeb Programming language support for balancing energy-efficiency and QoS
in mobile Web computing Abstraction Express QoS constraints Runtime Satisfy QoS specifications using energy saving techniques -
35 GreenWeb Programming language support for balancing energy-efficiency and QoS
in mobile Web computing Abstraction Express QoS constraints Runtime Satisfy QoS specifications using energy saving techniques Effect Significant energy savings -
37 GreenWeb: Language Extensions for Energy-Efficient Mobile Web Computing Yuhao
Zhu The University of Texas at Austin with Vijay Janapa Reddi PLDI 2016