Benchmarks of a Weather Forecasting Research Model on Superclusters Daniel B. Weber, Ph.D. Research Scientist CAPS/University of Oklahoma February 14,

Slides:



Advertisements
Similar presentations
SPEC High Performance Group (An Introduction)
Advertisements

Computer Organization, Bus Structure
Ali Baydoun1 Controllers (hard drive controllers).
Buffers & Spoolers J L Martin Think about it… All I/O is relatively slow. For most of us, input by typing is painfully slow. From the CPUs point.
Super Computers By Phuong Vo.
The AMD Athlon ™ Processor: Future Directions Fred Weber Vice President, Engineering Computation Products Group.
♦ Commodity processor with commodity inter- processor connection Clusters Pentium, Itanium, Opteron, Alpha GigE, Infiniband, Myrinet, Quadrics, SCI NEC.

Chapter 5: Server Hardware and Availability. Hardware Reliability and LAN The more reliable a component, the more expensive it is. Server hardware is.
Parallel I/O Performance Study Christian Chilan The HDF Group September 9, 2008SPEEDUP Workshop - HDF5 Tutorial1.
NPACI Panel on Clusters David E. Culler Computer Science Division University of California, Berkeley
CS 284a, 7 October 97Copyright (c) , John Thornley1 CS 284a Lecture Tuesday, 7 October 1997.
Question 7: How can you use this knowledge? Ghandi More Group A.
A Comparative Study of Network Protocols & Interconnect for Cluster Computing Performance Evaluation of Fast Ethernet, Gigabit Ethernet and Myrinet.
Parallel Programming on the SGI Origin2000 With thanks to Moshe Goldberg, TCC and Igor Zacharov SGI Taub Computer Center Technion Mar 2005 Anne Weill-Zrahia.
I/O Channels I/O devices getting more sophisticated e.g. 3D graphics cards CPU instructs I/O controller to do transfer I/O controller does entire transfer.
Lesson 12 – NETWORK SERVERS Distinguish between servers and workstations. Choose servers for Windows NT and Netware. Maintain and troubleshoot servers.
PARALLEL PROCESSING The NAS Parallel Benchmarks Daniel Gross Chen Haiout.
1 Microprocessor speeds Measure of system clock speed –How many electronic pulses the clock produces per second (clock frequency) –Usually expressed in.
Computer Systems Computer Performance.
Gordon: Using Flash Memory to Build Fast, Power-efficient Clusters for Data-intensive Applications A. Caulfield, L. Grupp, S. Swanson, UCSD, ASPLOS’09.
CPP Staff - 30 CPP Staff - 30 FCIPT Staff - 35 IPR Staff IPR Staff ITER-India Staff ITER-India Staff Research Areas: 1.Studies.
Cluster Computers. Introduction Cluster computing –Standard PCs or workstations connected by a fast network –Good price/performance ratio –Exploit existing.
Chapter 2 Computer Clusters Lecture 2.1 Overview.
Compaq - Indiana University Visit IU’s Compaq Parallel PC Cluster.
Chapter 2 Chapter 2: Planning for Server Hardware.
Computer Organization & Assembly Language
Computers Central Processor Unit. Basic Computer System MAIN MEMORY ALUCNTL..... BUS CONTROLLER Processor I/O moduleInterconnections BUS Memory.
Higher Computing Computer Systems S. McCrossan 1 Higher Grade Computing Studies 3. Computer Performance Measures of Processor Speed When comparing one.
Best Western Green Bay CHEMS 2013 SYSTEM ARCHITECTURE.
Adaptation of the Advanced Regional Prediction System to the Environmental Hydrology Workbench (Web-ARPS) Adaptation of the Advanced Regional Prediction.
Cluster Computers. Introduction Cluster computing –Standard PCs or workstations connected by a fast network –Good price/performance ratio –Exploit existing.
MCC website: ©Board of Trustees University of Illinois Research Objectives: Using game consoles as a platform for molecular modeling.
1 Selecting LAN server (Week 3, Monday 9/8/2003) © Abdou Illia, Fall 2003.
Chapter 2 Chapter 2: Planning for Server Hardware.
HYDRA: Using Windows Desktop Systems in Distributed Parallel Computing Arvind Gopu, Douglas Grover, David Hart, Richard Repasky, Joseph Rinkovsky, Steve.
Remote Direct Memory Access (RDMA) over IP PFLDNet 2003, Geneva Stephen Bailey, Sandburst Corp., Allyn Romanow, Cisco Systems,
Computer Architecture Lecture 3 Cache Memory. Characteristics Location Capacity Unit of transfer Access method Performance Physical type Physical characteristics.
Advanced Computer Architecture Cache Memory 1. Characteristics of Memory Systems 2.
SLAC Site Report Chuck Boeheim Assistant Director, SLAC Computing Services.
10/22/2002Bernd Panzer-Steindel, CERN/IT1 Data Challenges and Fabric Architecture.
Quick Introduction to NorduGrid Oxana Smirnova 4 th Nordic LHC Workshop November 23, 2001, Stockholm.
CHEP04 Performance Analysis of Cluster File System on Linux Yaodong CHENG IHEP, CAS
Easy Deployment of the WRF Model on Heterogeneous PC Systems Braden Ward and Shing Yoh Union, New Jersey.
Computing Environment The computing environment rapidly evolving ‑ you need to know not only the methods, but also How and when to apply them, Which computers.
Benchmarks of a Weather Forecasting Research Model Daniel B. Weber, Ph.D. Research Scientist CAPS/University of Oklahoma ****CONFIDENTIAL**** August 3,
KAIS T Computer Architecture Lab. Div. of CS, Dept. of EECS KAIST CS492 Lab Summary.
CASPUR Site Report Andrei Maslennikov Lead - Systems Rome, April 2006.
Coupling Facility. The S/390 Coupling Facility (CF), the key component of the Parallel Sysplex cluster, enables multisystem coordination and datasharing.
Installation of Storage Foundation for Windows High Availability 5.1 SP2 1 Daniel Schnack Principle Technical Support Engineer.
By Chi-Chang Chen.  Cluster computing is a technique of linking two or more computers into a network (usually through a local area network) in order.
WRF Software Development and Performance John Michalakes, NCAR NCAR: W. Skamarock, J. Dudhia, D. Gill, A. Bourgeois, W. Wang, C. Deluca, R. Loft NOAA/NCEP:
PROOF Benchmark on Different Hardware Configurations 1 11/29/2007 Neng Xu, University of Wisconsin-Madison Mengmeng Chen, Annabelle Leung, Bruce Mellado,
3/12/2013Computer Engg, IIT(BHU)1 PARALLEL COMPUTERS- 2.
Computer Hardware – Part 2 Basic Components V.T. Raja, Ph.D., Information Management Oregon State University.
Cluster Computers. Introduction Cluster computing –Standard PCs or workstations connected by a fast network –Good price/performance ratio –Exploit existing.
Computer Performance. Hard Drive - HDD Stores your files, programs, and information. If it gets full, you can’t save any more. Measured in bytes (KB,
Computer Hardware1 Hardware Types There is a great range of hardware, all of which is called computers. Bank Computers Supercomputers Minicomputers Microcomputers.
29/04/2008ALICE-FAIR Computing Meeting1 Resulting Figures of Performance Tests on I/O Intensive ALICE Analysis Jobs.
Community Grids Laboratory
Administration Tools Cluster.exe is a command line tool that you can use for scripting or remote administration through slow WAN links. Cluadmin.exe is.
Get more done with Windows 10 Pro for Workstations
Lab A: Installing and Configuring the Network Load Balancing Driver
Chapter 16: Distributed System Structures
File Transfer Protocol
By Brandon, Ben, and Lee Parallel Computing.
Computer type and application
Spooling Describe spooling, explaining why it is used.
Chapter 2: Planning for Server Hardware
Optimizing MPI collectives for SMP clusters
Presentation transcript:

Benchmarks of a Weather Forecasting Research Model on Superclusters Daniel B. Weber, Ph.D. Research Scientist CAPS/University of Oklahoma February 14, 2001

LL=UNM/Los Lobos 512 PIII/733 IBM RR=UNM/Road Runner 128 PII/450 AltaCluster NT=NCSA 128 PIII/550 NT Cluster TCS=Terascale Computing System/PSC 256 Alpha EV-67

INTEL Benchmark Summary n 20% increase in compute time for 2proc/node configuration on Intel Based systems due to bus competition n File system very slow on Intel based systems without fiber channel n File system is a weak link (UNM-LL) –5.5mb/sec sustained for 480 2proc/node tests writing 2.1mb files from 8 separate processors simultaneously –passing through linux file server not r6000

ALPHA Benchmark Summary n Alpha EV-67 (TCS) is 5 times faster computationally than the INTEL PIII/733 n Alpha (TCS) file system is very slow at times, need to look at the configuration, shows potential for very fast transfer rates n MPI overhead for a 256 processor TCS job is on the order of 15%, very good network performance.