Storage-aware Smartphone Energy Savings

458 views

Published on

UbiComp 2013 presentation by David T. Nguyen
College of William and Mary, USA

Published in: Education, Technology, Business
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
458
On SlideShare
0
From Embeds
0
Number of Embeds
7
Actions
Shares
0
Downloads
14
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Storage-aware Smartphone Energy Savings

  1. 1. Storage-aware Smartphone Energy Savings David T. Nguyen, Gang Zhou, Xin Qi, Ge Peng, Jianing Zhao,Tommy Nguyen, Duy Le
  2. 2. LIFE IN MOBILE ERA.. 1,038,000,000 SMARTPHONE USERS WORLDWIDE [IBTIMES] 27% INCREASED # SMARTPHONES SOLD ANNUALLY [IDC] Figure Courtesy: http://www.ideas4ios.com David T. Nguyen 2
  3. 3. SMARTPHONES EVERYWHERE! 75% AMERICANS USE THEM IN BATHROOMS [CBSNEWS] 50% USERS UNDER 25 USE THEM WHILE EATING [WILSON] Figure Courtesy: http://bublews.com David T. Nguyen 3
  4. 4. SMARTPHONE APPS DO EVERYTHING! 850,000 APPS IN APPLE STORE 05/13 [APPLE] 800,000 APPS IN GOOGLE PLAY 05/13 [CANALYS] 145,000 APPS IN WINDOWS STORE 05/13 [CANALYS] 120,000 APPS IN BLACKBERRY WORLD 05/13 [CANALYS] Figure Courtesy: http://aptito.com David T. Nguyen 4
  5. 5. Still BIG Problem David T. Nguyen 5 Figure Courtesy: http://cdn.cultofmac.com
  6. 6. Smartphone Dislikes David T. Nguyen 6 0% 10% 20% 30% 40% Call Quality Contact List Import Excessive Dropped Calls 3G Quality Screen Size 4G Capability Battery Life Source: ChangeWave
  7. 7. Outline Introduction Background Experimental Study SmartStorage Design Evaluation David T. Nguyen 7
  8. 8. Introduction  Researching energy consumption essential  What has been done ◦ Performance bottleneck in storage [Kim et al., FAST ‘12] ◦ No direct study of storage – energy consumption correlation David T. Nguyen 8
  9. 9. Introduction  Research questions ◦ How does storage affect smartphone power efficiency? ◦ How to optimize storage to save energy?  We propose SmartStorage ◦ Tracks smartphone I/O pattern ◦ Dynamically configures optimal storage parameters to save energy David T. Nguyen 9
  10. 10. Outline Introduction Background Experimental Study SmartStorage Design Evaluation David T. Nguyen 10
  11. 11. I/O Path David T. Nguyen 11 Red: Nexus One default static configurations
  12. 12. Outline Introduction Background Experimental Study SmartStorage Design Evaluation David T. Nguyen 12
  13. 13. Approach  Investigate impact of different storage configurations on power levels 1. Run series of benchmarks under default configurations 2. Repeat benchmarks under different configurations 3. Compare energy consumptions David T. Nguyen 13
  14. 14. Setup  Rooted smartphones: Nexus One, Nexus 4  8 benchmarks  Monsoon Power Monitor David T. Nguyen 14
  15. 15. Power Consumption: Default Config. (Queue Depth 128 / Write-back cache) David T. Nguyen 15  Different algorithms - different power levels  No algorithm optimal for all benchmarks  Changing algorithms may save energy
  16. 16. Power Consumption: Queue Depth 4 David T. Nguyen 16  Shorter queue depth saves energy in most cases  Not storage intensive benchmarks consume more power due to overhead of smaller queue
  17. 17. Power Consumption: Write-through Cache David T. Nguyen 17  Consumes less power  But requires rebuilding kernel  More details in paper…
  18. 18. Optimal Configurations  Run benchmarks with all combinations of scheduling algorithms and queue depths David T. Nguyen 18 Benchmark Optimal Conf. Power Savings Reads/s Writes/s AnTuTu Deadline/4 40% 1108 1395 CF-Bench CFQ/4 27% 104 1298 GLBenchmark Deadline/4 27% 253 51 BrowserMark CFQ/4 29% 185 115 AndroBench Noop/128 32% 2260 104 Quadrant BFQ/4 43% 301 400 Smartbench BFQ/128 0 26 2 Vellamo BFQ/128 0 9 1
  19. 19. Outline Introduction Background Experimental Study SmartStorage Design Evaluation David T. Nguyen 19
  20. 20. Big Idea Track phone’s run-time I/O pattern Match phone’s pattern with pattern from benchmark table Dynamically configure parameters with optimal savings David T. Nguyen 20
  21. 21. SmartStorage Architecture David T. Nguyen 21
  22. 22. GUI David T. Nguyen 22
  23. 23. I/O Pattern Matching  Compare phone’s I/O pattern with patterns from benchmark table  Matching feature: #𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 /𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 #𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 /𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠  If phone’s rate of reads/writes per second close to a benchmark from table ◦ match is found  Else ◦ no match David T. Nguyen 23
  24. 24. Outline Introduction Background Experimental Study SmartStorage Design Evaluation David T. Nguyen 24
  25. 25. Energy Savings: Nexus One David T. Nguyen 25  3 apps w/ no savings – same default and optimal configs (BFQ/128)
  26. 26. Energy Savings: Nexus 4 David T. Nguyen 26  Lower savings due to default CFQ scheduler  Average savings of 28.8%
  27. 27. Discussion  Savings of whole phone  Savings come from optimizations on I/O path (not flash only)  How optimizations affect CPU and other subsystems still unknown  Cost: 3% app delay David T. Nguyen 27
  28. 28. Real-time Power David T. Nguyen 28
  29. 29. Real-time Power David T. Nguyen 29 Power drop  Power drops after around 2 minutes (loading + I/O pattern recalculation)
  30. 30. Conclusions  Presented study on how storage parameters impact power levels  Introduced SmartStorage to save energy ◦ Matches current I/O pattern to known pattern from benchmarks ◦ Dynamically tunes parameters  Evaluation on top 20 apps shows on average 28.8% energy savings David T. Nguyen 30
  31. 31. Future Work  Energy savings with different caching policies / file systems / queue depths  Matching using machine learning  Adaptive I/O pattern recalculation  Root reasons of energy savings David T. Nguyen 31
  32. 32. Current Project Status  U.S. patent filed 12/2012  Please forward licensing inquiries to ◦ William & Mary TechnologyTransfer Office David T. Nguyen 32
  33. 33. PROJECT WEBSITE SmartStorage.us David T. Nguyen 33
  34. 34. THANKYOU! David T. Nguyen 34

×