CSE 591: Energy-Efficient Computing Lecture 3 SPEED: processor Anshul Gandhi 347, CS building

Slides:



Advertisements
Similar presentations
Energy-efficient Task Scheduling in Heterogeneous Environment 2013/10/25.
Advertisements

Zhou Peng, Zuo Decheng, Zhou Haiying Harbin Institute of Technology 1.
S YSTEM -W IDE E NERGY M ANAGEMENT FOR R EAL -T IME T ASKS : L OWER B OUND AND A PPROXIMATION Xiliang Zhong and Cheng-Zhong Xu ICCAD 2006, ACM Trans. on.
International Symposium on Low Power Electronics and Design Energy-Efficient Non-Minimal Path On-chip Interconnection Network for Heterogeneous Systems.
CSE 691: Energy-Efficient Computing Lecture 20 review Anshul Gandhi 1307, CS building
1 MemScale: Active Low-Power Modes for Main Memory Qingyuan Deng, David Meisner*, Luiz Ramos, Thomas F. Wenisch*, and Ricardo Bianchini Rutgers University.
A Cyber-Physical Systems Approach to Energy Management in Data Centers Presented by Chen He Adopted form the paper authors.
CSE 691: Energy-Efficient Computing Lecture 4 SCALING: stateless vs. stateful Anshul Gandhi 1307, CS building
Power Cost Reduction in Distributed Data Centers Yuan Yao University of Southern California 1 Joint work: Longbo Huang, Abhishek Sharma, LeanaGolubchik.
1 Introduction Background: CS 3810 or equivalent, based on Hennessy and Patterson’s Computer Organization and Design Text for CS/EE 6810: Hennessy and.
SLA-aware Virtual Resource Management for Cloud Infrastructures
CSE 691: Energy-Efficient Computing Lecture 3 SLEEP: full-system Anshul Gandhi 1307, CS building
Energy, Energy, Energy  Worldwide efforts to reduce energy consumption  People can conserve. Large percentage savings possible, but each individual has.
Synergy.cs.vt.edu Power and Performance Characterization of Computational Kernels on the GPU Yang Jiao, Heshan Lin, Pavan Balaji (ANL), Wu-chun Feng.
Task Dependence in Scheduling and Load Balancing Prof. Adam Meyerson UCLA.
Power-Aware SoC Test Optimization through Dynamic Voltage and Frequency Scaling Vijay Sheshadri, Vishwani D. Agrawal, Prathima Agrawal Dept. of Electrical.
Green IT and Data Centers Darshan R. Kapadia Gregor von Laszewski 1.
Folklore Confirmed: Compiling for Speed = Compiling for Energy Tomofumi Yuki INRIA, Rennes Sanjay Rajopadhye Colorado State University 1.
XI HE Computing and Information Science Rochester Institute of Technology Rochester, NY USA Rochester Institute of Technology Service.
OPTIMAL SERVER PROVISIONING AND FREQUENCY ADJUSTMENT IN SERVER CLUSTERS Presented by: Xinying Zheng 09/13/ XINYING ZHENG, YU CAI MICHIGAN TECHNOLOGICAL.
Lecture 03: Fundamentals of Computer Design - Trends and Performance Kai Bu
Cloud Computing Energy efficient cloud computing Keke Chen.
DENS: Data Center Energy-Efficient Network-Aware Scheduling
Last Time Performance Analysis It’s all relative
CSE 691: Energy-Efficient Computing Lecture 6 SHARING: distributed vs. local Anshul Gandhi 1307, CS building
CSE 691: Energy-Efficient Computing Lecture 7 SMARTS: custom-made systems Anshul Gandhi 1307, CS building
Temperature Aware Load Balancing For Parallel Applications Osman Sarood Parallel Programming Lab (PPL) University of Illinois Urbana Champaign.
Critical Power Slope Understanding the Runtime Effects of Frequency Scaling Akihiko Miyoshi, Charles Lefurgy, Eric Van Hensbergen Ram Rajamony Raj Rajkumar.
1 CS/EE 6810: Computer Architecture Class format:  Most lectures on YouTube *BEFORE* class  Use class time for discussions, clarifications, problem-solving,
Challenges towards Elastic Power Management in Internet Data Center.
[Tim Shattuck, 2006][1] Performance / Watt: The New Server Focus Improving Performance / Watt For Modern Processors Tim Shattuck April 19, 2006 From the.
Eneryg Efficiency for MapReduce Workloads: An Indepth Study Boliang Feng Renmin University of China Dec 19.
Abhilash Thekkilakattil, Radu Dobrin, Sasikumar Punnekkat Mälardalen Real-time Research Center, Mälardalen University Västerås, Sweden Towards Preemption.
Power-Aware Scheduling of Virtual Machines in DVFS-enabled Clusters
CSE 691: Energy-Efficient Computing Lecture 1: Intro and Logistics Anshul Gandhi 1307, CS building
Copyright © 2011, Performance Evaluation of a Green Scheduling Algorithm for Energy Savings in Cloud Computing Truong Vinh Truong Duy; Sato,
Towards Dynamic Green-Sizing for Database Servers Mustafa Korkmaz, Alexey Karyakin, Martin Karsten, Kenneth Salem University of Waterloo.
Power and Control in Networked Sensors E. Jason Riedy and Robert Szewczyk Presenter: Fayun Luo.
An Energy-efficient Task Scheduler for Multi-core Platforms with per-core DVFS Based on Task Characteristics Ching-Chi Lin Institute of Information Science,
Critical Power Slope: Understanding the Runtime Effects of Frequency Scaling Akihiko Miyoshi †,Charles Lefurgy ‡, Eric Van Hensbergen ‡, Ram Rajamony ‡,
XI HE Computing and Information Science Rochester Institute of Technology Rochester, NY USA Rochester Institute of Technology Service.
Multimedia Computing and Networking Jan Reduced Energy Decoding of MPEG Streams Malena Mesarina, HP Labs/UCLA CS Dept Yoshio Turner, HP Labs.
Accounting for Load Variation in Energy-Efficient Data Centers
Xi He Golisano College of Computing and Information Sciences Rochester Institute of Technology Rochester, NY THERMAL-AWARE RESOURCE.
Workload Clustering for Increasing Energy Savings on Embedded MPSoCs S. H. K. Narayanan, O. Ozturk, M. Kandemir, M. Karakoy.
1 Lecture 2: Metrics to Evaluate Systems Topics: Metrics: power, reliability, cost, benchmark suites, performance equation, summarizing performance with.
CSE 591: Energy-Efficient Computing Lecture 1: Intro and Logistics Anshul Gandhi 347, New CS building
CSE 591: Energy-Efficient Computing Lecture 4 SLEEP: full-system Anshul Gandhi 347, CS building
CSE 591: Energy-Efficient Computing Lecture 8 SOURCE: renewables Anshul Gandhi 347, CS building
Determining Optimal Processor Speeds for Periodic Real-Time Tasks with Different Power Characteristics H. Aydın, R. Melhem, D. Mossé, P.M. Alvarez University.
ECE 692 Power-Aware Computer Systems Final Review Prof. Xiaorui Wang.
Power Capping Via Forced Idleness ANSHUL GANDHI Carnegie Mellon Univ. 1.
CSE 340 Computer Architecture Summer 2016 Understanding Performance.
CSE 591: Energy-Efficient Computing Lecture 13 SLEEP: sensors
Energy Aware Network Operations
Anshul Gandhi 347, CS building
Anshul Gandhi 347, CS building
Andrea Acquaviva, Luca Benini, Bruno Riccò
CSE 591: Energy-Efficient Computing Lecture 17 SCALING: survey
CSE 591: Energy-Efficient Computing Lecture 20 SPEED: disks
CSE 591: Energy-Efficient Computing Lecture 15 SCALING: storage
CSE 591: Energy-Efficient Computing Lecture 10 SLEEP: network
CSE 591: Energy-Efficient Computing Lecture 19 SPEED: memory
Optimal Power Allocation in Server Farms
Hui Chen, Shinan Wang and Weisong Shi Wayne State University
CSE 591: Energy-Efficient Computing Lecture 12 SLEEP: memory
CSE 591: Energy-Efficient Computing Lecture 14 SCALING: setup time
CSE 591: Energy-Efficient Computing Lecture 9 SLEEP: processor
CSE 531: Performance Analysis of Systems Lecture 4: DTMC
CSE 591: Energy-Efficient Computing Lecture 18 SPEED: power
Presentation transcript:

CSE 591: Energy-Efficient Computing Lecture 3 SPEED: processor Anshul Gandhi 347, CS building

opt_allocation paper

U.S. Data Center Energy Consumption 3 $ 8.4 billion kWh (in billions)  120 billion kWh 12 billion kWh 50 billion kWh Source: EPA report to Congress on Server and Data Center Energy Efficiency,2007

4 P Get the best performance from the power, P, that we have. Goal Data Center

5 P P1P1 P2P2 P3P3 Goal How to split P to minimize mean response time? Right answer can improve performance by up to 5X Constraint: P ≥ P 1 + P 2 + P 3

Our Experimental Results DFS: Dynamic Frequency Scaling Power (Watts) DFS 6 Frequency (GHz) (server speed) How power affects server speed for a single server “linear” P = system power NOT processor power

Our Experimental Results Power (Watts) DFS 7 Frequency (GHz) How power affects server speed for a single server Power (Watts) Frequency (GHz) Power (Watts) Frequency (GHz) DVFS +DFS Power (Watts) DFS Frequency (GHz) Power (Watts) Frequency (GHz) Power (Watts) Frequency (GHz) DVFS +DFS “LINPACK” CPU BOUND “STREAM” MEM BOUND

Power Allocation Results 8 CPU bound “LINPACK” Memory bound “STREAM” DFS DVFS DVFS+DFS Power (Watts) DFS Frequency (GHz)

Power Allocation Results 9 CPU bound “LINPACK” Memory bound “STREAM” DFS DVFS DVFS+DFS Power (Watts) Frequency (GHz) DVFS

Power Allocation Results 10 CPU bound “LINPACK” Memory bound “STREAM” DFS DVFS DVFS+DFS Power (Watts) Frequency (GHz) DVFS +DFS

Power Allocation Results 11 CPU bound “LINPACK” Memory bound “STREAM” DFS DVFS DVFS+DFS DFS DVFS Arrival rate (jobs/sec) Mean Resp. Time (sec) Arrival rate (jobs/sec)

Conclusions: How to allocate power optimally 12 Speed Scaling? Arrival Rate? Linear, SteepLinear, FlatCubic Arrival Rate? PowMax PowMin High LowHigh Low High Low PowMax PowMed