Presentation is loading. Please wait.

Presentation is loading. Please wait.

MARS: Adaptive Remote Execution Scheduler for Multithreaded Mobile Devices Asaf Cidon*, Tomer M. London*, Sachin Katti, Christos Kozyrakis, Mendel Rosenblum.

Similar presentations


Presentation on theme: "MARS: Adaptive Remote Execution Scheduler for Multithreaded Mobile Devices Asaf Cidon*, Tomer M. London*, Sachin Katti, Christos Kozyrakis, Mendel Rosenblum."— Presentation transcript:

1 MARS: Adaptive Remote Execution Scheduler for Multithreaded Mobile Devices Asaf Cidon*, Tomer M. London*, Sachin Katti, Christos Kozyrakis, Mendel Rosenblum *Equal contributors Stanford University

2 New Class of Mobile Applications October 23, 2011Slide 2 Augmented Reality Computer Vision Motion Sensing

3 Mobile Client Trends Mobile CPU performance increasing – Hitting energy wall Can we improve performance and reduce energy consumption? Opportunity: network bandwidth increase utilize the cloud Slide 3October 23, 2011 Maximum Bandwidth (Mb/s)

4 Static Client-Server Partitioning Doesnt Work Dynamic resources: – Network bandwidth and latency – Available CPU, memory Same code, different platforms: – Smartphones (single-core, multi-core) – Tablets October 23, 2011Slide 4

5 MARS: Adaptive Remote Execution Opportunistically offload computations to remote server – Enhance computational capabilities – Decrease energy consumption Make dynamic decisions – Adapt to network and CPU variability October 23, 2011Slide 5 Data Center Mobile Device

6 Agenda 1.Design of MARS 2.Simulator Results and Analysis 3.Conclusions October 23, 2011Slide 6

7 Existing Remote Execution Systems October 23, 2011Slide 7 The Unit of Remote Execution Target of Performance Optimization RPC VM Single-thread application Multi-threaded application System CloneCloud [Kirsch et al., 11] Cloudlets [Satyanarayanan et al., 09] MAUI [Cuervo et al. 10] Chroma [Balan et al. 03] Odessa [Ra et al. 11] MARS Cloud-on- Chip

8 Previous Systems: Application Partitioning October 23, 2011Slide 8 RPC 1 Process 1 RPC 2 Process 1 RPC 3 Process 1 RPC 4 Process 1 RPC 5 Process 1 Local ExecutionRemote Execution RPC 2 Process 3 RPC 1 Process 3 RPC 2 Process 1 RPC 1 Process 2 RPC 1 Process 1 RPC Queue Local Cores Remote Cores MARS Cloud-on-Chip: System Scheduling

9 Greedy Algorithm Slide 9October 23, 2011 Higher POR: better performance gain from offloading Higher EOR: better energy saving from offloading

10 Remote Server Local Core Controller Algorithm Slide 10 October 23, 2011 Priority Queue, sorted by Performance Offload Rank (POR) Available EOR LocalRemoteBoth Check EOR Threshold G (Greediness) trades-off utilization and energy efficiency RPC 2 (POR 0.4) RPC 4 (POR 1.3) RPC 6 (POR 1.8) RPC 5 (POR 1.9) RPC 3 (POR 2.5) RPC 6 (POR 1.8)

11 Agenda 1.Design of MARS 2.Simulator Results and Analysis 3.Conclusions October 23, 2011Slide 11

12 Remote Execution Applications Detection Recognition Pic Barcode Rendering Pic Slide 12 Barcode Rendering Pic Barcode Rendering Pic Detection Recognition Pic Detection Recognition Pic Augmented RealityFace Recognition

13 Simulator Methodology Trace-driven simulation Clients: – Nokia N900 (single core) – NVIDIA Tegra 250 (multicore) Server: – Amazon EC2 Opteron 2007 Networks: – Outdoors Wi-Fi – Indoors Wi-Fi – 3G Slide 13June 4, 2011

14 MARS vs. Static Policies Slide 14

15 Nokia N900 Power Consumption WiFi: Performance and energy are highly correlated 3G: trade-off performance and energy October 23, 2011Slide 15 Wi-Fi3G Idle Network Power1.31 Watts0.66 Watts Upload Network Power Watts2.36 Watts Download Network Power 1.39 Watts2.26 Watts Upload Network Power Overhead 10.51%72.03%

16 Same Application, Different Networks Slide 16

17 Remote Execution with Multicore Slide 17October 23, 2011

18 Agenda 1.Design of MARS 2.Simulator Results and Analysis 3.Conclusions October 23, 2011Slide 18

19 Conclusions 1.Cant always be greedy – Performance and energy trade-off 2.MARS is optimized for multiple parallel applications and cores 3.MARS Cloud-on-Chip: validation of system- level remote execution scheduling – 57% performance increase, 33% energy savings October 23, 2011Slide 19


Download ppt "MARS: Adaptive Remote Execution Scheduler for Multithreaded Mobile Devices Asaf Cidon*, Tomer M. London*, Sachin Katti, Christos Kozyrakis, Mendel Rosenblum."

Similar presentations


Ads by Google