Desktop Workload Characterization for CMP/SMT and Implications for Operating System Design Sven Bachthaler Fernando Belli Alexandra Fedorova Simon Fraser.

Slides:



Advertisements
Similar presentations
KAIST Computer Architecture Lab. The Effect of Multi-core on HPC Applications in Virtualized Systems Jaeung Han¹, Jeongseob Ahn¹, Changdae Kim¹, Youngjin.
Advertisements

Performance of Multithreaded Chip Multiprocessors and Implications for Operating System Design Hikmet Aras
Dynamic Load Balancing for VORPAL Viktor Przebinda Center for Integrated Plasma Studies.
S.Chechelnitskiy / SFU Simon Fraser Running CE and SE in a XEN virtualized environment S.Chechelnitskiy Simon Fraser University CHEP 2007 September 6 th.
Operating System CS105. Objectives Role of an operating system Manages resources – Memory – CPU – Secondary storage – I/O devices Memory CPU Hard Disk.
®® Microsoft Windows 7 Windows Tutorial 8 Connecting to Networks with Mobile Computing.
Copyright © 2005 Department of Computer Science CPSC 641 Winter PERFORMANCE EVALUATION Often in Computer Science you need to: – demonstrate that.
JProbe. 1. JProbe Use JProbe Profile –identify method and line level performance bottlenecks Use JProbe Memory Debugger –investigating memory leaks and.
Enter Date in Title Master Workload Management HBC Case Study IRMAC, January 2008 Shelley Perrior -DBA team lead.
Effectively Utilizing Global Cluster Memory for Large Data-Intensive Parallel Programs John Oleszkiewicz, Li Xiao, Yunhao Liu IEEE TRASACTION ON PARALLEL.
A Grid Resource Broker Supporting Advance Reservations and Benchmark- Based Resource Selection Erik Elmroth and Johan Tordsson Reporter : S.Y.Chen.
70-290: MCSE Guide to Managing a Microsoft Windows Server 2003 Environment Chapter 11: Monitoring Server Performance.
MCDST : Supporting Users and Troubleshooting a Microsoft Windows XP Operating System Chapter 6: Configure and Troubleshoot Local User and Group Accounts.
1 PERFORMANCE EVALUATION H Often in Computer Science you need to: – demonstrate that a new concept, technique, or algorithm is feasible –demonstrate that.
Job scheduling Queue discipline.
SyNAR: Systems Networking and Architecture Group Symbiotic Jobscheduling for a Simultaneous Multithreading Processor Presenter: Alexandra Fedorova Simon.
ABACUS: A Hardware-Based Software Profiler for Modern Processors Eric Matthews Lesley Shannon School of Engineering Science Sergey Blagodurov Sergey Zhuravlev.
Module 8: Monitoring SQL Server for Performance. Overview Why to Monitor SQL Server Performance Monitoring and Tuning Tools for Monitoring SQL Server.
Measuring zSeries System Performance Dr. Chu J. Jong School of Information Technology Illinois State University 06/11/2012 Sponsored in part by Deer &
How Significant Is the Effect of Faults Interaction on Coverage Based Fault Localizations? Xiaozhen Xue Advanced Empirical Software Testing Group Department.
SYNAR Systems Networking and Architecture Group Scheduling on Heterogeneous Multicore Processors Using Architectural Signatures Daniel Shelepov and Alexandra.
Department of Computer Science Mining Performance Data from Sampled Event Traces Bret Olszewski IBM Corporation – Austin, TX Ricardo Portillo, Diana Villa,
Unifying Primary Cache, Scratch, and Register File Memories in a Throughput Processor Mark Gebhart 1,2 Stephen W. Keckler 1,2 Brucek Khailany 2 Ronny Krashinsky.
A Virtual Instrumental Analysis Laboratory (VIAL) for Buffalo State and Open SUNY Jinseok Heo, Alexander Nazarenko, M. Scott Goodman, and Jamie Kime Department.
1 Wenguang WangRichard B. Bunt Department of Computer Science University of Saskatchewan November 14, 2000 Simulating DB2 Buffer Pool Management.
How to use Remote Desktop and Remote Support. What is remote desktop? Remotely control your computer from another office, from home, or while traveling.
Profiling Memory Subsystem Performance in an Advanced POWER Virtualization Environment The prominent role of the memory hierarchy as one of the major bottlenecks.
ACMSE’04, ALDepartment of Electrical and Computer Engineering - UAH Execution Characteristics of SPEC CPU2000 Benchmarks: Intel C++ vs. Microsoft VC++
Running a Scientific Experiment on the Grid Vilnius, 13 rd May, 2008 by Tomasz Szepieniec IFJ PAN & CYFRONET.
The Owner Share scheduler for a distributed system 2009 International Conference on Parallel Processing Workshops Reporter: 李長霖.
Srihari Makineni & Ravi Iyer Communications Technology Lab
OPERATING SYSTEMS Lecture 3: we will explore the role of the operating system in a computer Networks and Communication Department 1.
Autonomic scheduling of tasks from data parallel patterns to CPU/GPU core mixes Published in: High Performance Computing and Simulation (HPCS), 2013 International.
70-290: MCSE Guide to Managing a Microsoft Windows Server 2003 Environment, Enhanced Chapter 11: Monitoring Server Performance.
1 Oracle Enterprise Manager Slides from Dominic Gélinas CIS
Memory Performance Profiling via Sampled Performance Monitor Event Traces Diana Villa, Patricia J. Teller, and Jaime Acosta The University of Texas at.
Embedded System Lab. 정범종 A_DRM: Architecture-aware Distributed Resource Management of Virtualized Clusters H. Wang et al. VEE, 2015.
Compiler and Runtime Support for Enabling Generalized Reduction Computations on Heterogeneous Parallel Configurations Vignesh Ravi, Wenjing Ma, David Chiu.
Chapter 3 System Performance and Models Introduction A system is the part of the real world under study. Composed of a set of entities interacting.
Department of Computer Science MapReduce for the Cell B. E. Architecture Marc de Kruijf University of Wisconsin−Madison Advised by Professor Sankaralingam.
Design Issues of Prefetching Strategies for Heterogeneous Software DSM Author :Ssu-Hsuan Lu, Chien-Lung Chou, Kuang-Jui Wang, Hsiao-Hsi Wang, and Kuan-Ching.
Module 9 Planning and Implementing Monitoring and Maintenance.
Template This is a template to help, not constrain, you. Modify as appropriate. Move bullet points to additional slides as needed. Don’t cram onto a single.
Advanced Hardware/Software Optimization Techniques for Application Specific MCSoC m Yumiko Kimezawa Supervised by Prof. Ben Abderazek Adapted Systems.
A Two-phase Execution Engine of Reduce Tasks In Hadoop MapReduce XiaohongZhang*GuoweiWang* ZijingYang*YangDing School of Computer Science and Technology.
Sunpyo Hong, Hyesoon Kim
Department of Computer Science 6 th Annual Austin CAS Conference – 24 February 2005 Ricardo Portillo, Diana Villa, Patricia J. Teller The University of.
Software Engineering Prof. Dr. Bertrand Meyer March 2007 – June 2007 Chair of Software Engineering Lecture #20: Profiling NetBeans Profiler 6.0.
Concurrency Conclusion Andy Wang Operating Systems COP 4610 / CGS 5765.
Capacity Planning in a Virtual Environment Chris Chesley, Sr. Systems Engineer
1 Evaluation of Cooperative Web Caching with Web Polygraph Ping Du and Jaspal Subhlok Department of Computer Science University of Houston presented at.
LIOProf: Exposing Lustre File System Behavior for I/O Middleware
Spark on Entropy : A Reliable & Efficient Scheduler for Low-latency Parallel Jobs in Heterogeneous Cloud Huankai Chen PhD Student at University of Kent.
Department of Computer Science DCS COMSATS Institute of Information Technology Thread level parallelism of desktop applications Presented by Muhammad Kamran.
OPERATING SYSTEMS CS 3502 Fall 2017
Thesis Proposal Title Student Name Registration No. Supervisor Name
CPU Scheduling.
Collaborative Offloading for Distributed Mobile-Cloud Apps
Computer Science I CSC 135.
Department of Computer Science University of California, Santa Barbara
Miss rate versus (period, slice)
Future Year Scheduling
Operating Systems.
Computer Systems Performance Evaluation
Thesis Proposal Title Student Name Registration No. Supervisor Name
EE 472 – Embedded Systems Dr. Shwetak Patel.
Computer Systems Performance Evaluation
[Most of the details about queues are left for to read about and work out in Lab 6.] Def. As a data structure, a queue is an ordered collection of data.
Concurrency Conclusion
Concurrency Conclusion
Presentation transcript:

Desktop Workload Characterization for CMP/SMT and Implications for Operating System Design Sven Bachthaler Fernando Belli Alexandra Fedorova Simon Fraser University Canada

Objectives  Advanced scheduling algorithms for desktop systems?  Data collection from live systems

Motivation  First study for desktop systems (restricted to Windows XP)  Should we address parallelism in periods of activity?

Approach  Metric for parallelism  Ready queue length  Characterization of parallelism  Zero parallelism(no threads waiting)  Low parallelism (1-2 threads waiting)  High parallelism(>2 threads waiting)

Outline  Methodology and Data Collection  Results  Conclusions  Future Work

Methodology  Collect data from three groups  20 university lab computers  10 university staff computers  12 home computers

Methodology  Local and remote data collection  Remote data collection  For university computers  Less overhead  No user interaction necessary  Local data collection for home PCs

Tools  Performance Monitor  PsList  PsInfo

Data Collection  Collected every 15 seconds:  Ready queue length  Number of running processes  Number of running threads  Available main memory  Percentage of time when CPU is busy

Results  Presenting the results  Each slide for specific hardware  Several computers grouped according to hardware configuration

Results  University lab computers…

Results  Three groups of lab computers

Results  Three lab computers

Results  University staff computers…

Results  Single staff computer

Results  Six staff computers

Results  Home computers…

Results  Home computers without CMP/SMT

Results  Three home computers with CMP/SMT

Results  Special case…

Results  Staff computer

Conclusion  Low parallelism for a significant number of analyzed workloads  Not too much benefit from performance-optimizing scheduling algorithms

Future Work  Expand data collection to gain statistical significance  Investigate better ways for local data collection

Acknowledgements  We want to thank the department of Computing Science at SFU  Special thanks to the volunteers for the data collection  Thank you!