Pathway to Petaflops A vendor contribution Philippe Trautmann Business Development Manager HPC & Grid Global Education, Government & Healthcare.

Slides:



Advertisements
Similar presentations
Computing Infrastructure
Advertisements

PowerEdge T20 Customer Presentation. Product overview Customer benefits Use cases Summary PowerEdge T20 Overview 2 PowerEdge T20 mini tower server.
Oracle Exadata for SAP.
1 GridTorrent Framework: A High-performance Data Transfer and Data Sharing Framework for Scientific Computing.
IBM 1350 Cluster Expansion Doug Johnson Senior Systems Developer.
CURRENT AND FUTURE HPC SOLUTIONS. T-PLATFORMS  Russia’s leading developer of turn-key solutions for supercomputing  Privately owned  140+ employees.
SAN DIEGO SUPERCOMPUTER CENTER at the UNIVERSITY OF CALIFORNIA; SAN DIEGO SDSC RP Update October 21, 2010.
LinkSCEEM-2: A computational resource for the development of Computational Sciences in the Eastern Mediterranean Mostafa Zoubi SESAME SESAME – LinkSCEEM.
Linux Clustering A way to supercomputing. What is Cluster? A group of individual computers bundled together using hardware and software in order to make.
NWfs A ubiquitous, scalable content management system with grid enabled cross site data replication and active storage. R. Scott Studham.
Sun FIRE Jani Raitavuo Niko Ronkainen. Sun FIRE 15K Most powerful and scalable Up to 106 processors, 576 GB memory and 250 TB online disk storage Fireplane.
Data Storage Willis Kim 14 May Types of storages Direct Attached Storage – storage hardware that connects to a single server Direct Attached Storage.
Virtual Network Servers. What is a Server? 1. A software application that provides a specific one or more services to other computers  Example: Apache.
Real Parallel Computers. Modular data centers Background Information Recent trends in the marketplace of high performance computing Strohmaier, Dongarra,
Cluster computing facility for CMS simulation work at NPD-BARC Raman Sehgal.
KYLIN-I 麒麟一号 High-Performance Computing Cluster Institute for Fusion Theory and Simulation, Zhejiang University
Database Services for Physics at CERN with Oracle 10g RAC HEPiX - April 4th 2006, Rome Luca Canali, CERN.
© Copyright 2010 Hewlett-Packard Development Company, L.P. 1 HP + DDN = A WINNING PARTNERSHIP Systems architected by HP and DDN Full storage hardware and.
Computing/Tier 3 Status at Panjab S. Gautam, V. Bhatnagar India-CMS Meeting, Sept 27-28, 2007 Delhi University, Delhi Centre of Advanced Study in Physics,
Corporate Partner Overview and Update September 27, 2007 Gary Crane SURA Director IT Initiatives.
High Performance Computing G Burton – ICG – Oct12 – v1.1 1.
US ATLAS Western Tier 2 Status and Plan Wei Yang ATLAS Physics Analysis Retreat SLAC March 5, 2007.
Voltaire The Grid Backbone™ InfiniBand CERN Seminar Asaf Somekh VP Strategic Alliances June 2006.
9/16/2000Ian Bird/JLAB1 Planning for JLAB Computational Resources Ian Bird.
Maximizing The Compute Power With Mellanox InfiniBand Connectivity Gilad Shainer Wolfram Technology Conference 2006.
UTA Site Report Jae Yu UTA Site Report 4 th DOSAR Workshop Iowa State University Apr. 5 – 6, 2007 Jae Yu Univ. of Texas, Arlington.
3. April 2006Bernd Panzer-Steindel, CERN/IT1 HEPIX 2006 CPU technology session some ‘random walk’
Nov 1, 2000Site report DESY1 DESY Site Report Wolfgang Friebel DESY Nov 1, 2000 HEPiX Fall
Rensselaer Why not change the world? Rensselaer Why not change the world? 1.
Hardware Trends. Contents Memory Hard Disks Processors Network Accessories Future.
Scientific Computing Experimental Physics Lattice QCD Sandy Philpott May 20, 2011 IT Internal Review 12GeV Readiness.
Taking the Complexity out of Cluster Computing Vendor Update HPC User Forum Arend Dittmer Director Product Management HPC April,
Network Tests at CHEP K. Kwon, D. Han, K. Cho, J.S. Suh, D. Son Center for High Energy Physics, KNU, Korea H. Park Supercomputing Center, KISTI, Korea.
21 st October 2002BaBar Computing – Stephen J. Gowdy 1 Of 25 BaBar Computing Stephen J. Gowdy BaBar Computing Coordinator SLAC 21 st October 2002 Second.
6/26/01High Throughput Linux Clustering at Fermilab--S. Timm 1 High Throughput Linux Clustering at Fermilab Steven C. Timm--Fermilab.
SoCal Infrastructure OptIPuter Southern California Network Infrastructure Philip Papadopoulos OptIPuter Co-PI University of California, San Diego Program.
Massive Supercomputing Coping with Heterogeneity of Modern Accelerators Toshio Endo and Satoshi Matsuoka Tokyo Institute of Technology, Japan.
Experience with the Thumper Wei Yang Stanford Linear Accelerator Center May 27-28, 2008 US ATLAS Tier 2/3 workshop University of Michigan, Ann Arbor.
JLab Scientific Computing: Theory HPC & Experimental Physics Thomas Jefferson National Accelerator Facility Newport News, VA Sandy Philpott.
RAL Site Report Andrew Sansum e-Science Centre, CCLRC-RAL HEPiX May 2004.
Infiniband in EDA (Chip Design) Glenn Newell Sr. Staff IT Architect Synopsys.
ITEP computing center and plans for supercomputing Plans for Tier 1 for FAIR (GSI) in ITEP  8000 cores in 3 years, in this year  Distributed.
HEPix April 2006 NIKHEF site report What’s new at NIKHEF’s infrastructure and Ramping up the LCG tier-1 Wim Heubers / NIKHEF (+SARA)
Next Generation Operating Systems Zeljko Susnjar, Cisco CTG June 2015.
CERN IT Department CH-1211 Genève 23 Switzerland t Frédéric Hemmer IT Department Head - CERN 23 rd August 2010 Status of LHC Computing from.
ATLAS Tier 1 at BNL Overview Bruce G. Gibbard Grid Deployment Board BNL 5-6 September 2006.
IDC HPC User Forum April 14 th, 2008 A P P R O I N T E R N A T I O N A L I N C Steve Lyness Vice President, HPC Solutions Engineering
CERN Computer Centre Tier SC4 Planning FZK October 20 th 2005 CERN.ch.
Infrastructure for Data Warehouses. Basics Of Data Access Data Store Machine Memory Buffer Memory Cache Data Store Buffer Bus Structure.
Technology Summary John Gordon. Talks University Multidisciplinary Scientific Computing: Experience and Plans - Alan Tackett (Vanderbilt University) PASTA.
Tier-2 storage A hardware view. HEP Storage dCache –needs feed and care although setup is now easier. DPM –easier to deploy xrootd (as system) is also.
PROOF tests at BNL Sergey Panitkin, Robert Petkus, Ofer Rind BNL May 28, 2008 Ann Arbor, MI.
Presented by NCCS Hardware Jim Rogers Director of Operations National Center for Computational Sciences.
Computing Issues for the ATLAS SWT2. What is SWT2? SWT2 is the U.S. ATLAS Southwestern Tier 2 Consortium UTA is lead institution, along with University.
 System Requirements are the prerequisites needed in order for a software or any other resources to execute efficiently.  Most software defines two.
Tackling I/O Issues 1 David Race 16 March 2010.
Latest Improvements in the PROOF system Bleeding Edge Physics with Bleeding Edge Computing Fons Rademakers, Gerri Ganis, Jan Iwaszkiewicz CERN.
The Worldwide LHC Computing Grid Frédéric Hemmer IT Department Head Visit of INTEL ISEF CERN Special Award Winners 2012 Thursday, 21 st June 2012.
UTA Site Report Jae Yu UTA Site Report 7 th DOSAR Workshop Louisiana State University Apr. 2 – 3, 2009 Jae Yu Univ. of Texas, Arlington.
PetaCache: Data Access Unleashed Tofigh Azemoon, Jacek Becla, Chuck Boeheim, Andy Hanushevsky, David Leith, Randy Melen, Richard P. Mount, Teela Pulliam,
Dominique Boutigny December 12, 2006 CC-IN2P3 a Tier-1 for W-LCG 1 st Chinese – French Workshop on LHC Physics and associated Grid Computing IHEP - Beijing.
Voltaire and the CERN openlab collaborate on Grid technology project using InfiniBand May 27, 2004 Patrick Chevaux EMEA Business Development
Architecture of a platform for innovation and research Erik Deumens – University of Florida SC15 – Austin – Nov 17, 2015.
Peter Idoine Managing Director Oracle New Zealand Limited.
Jefferson Lab Site Report Sandy Philpott HEPiX Fall 07 Genome Sequencing Center Washington University at St. Louis.
Scientific Computing at Fermilab Lothar Bauerdick, Deputy Head Scientific Computing Division 1 of 7 10k slot tape robots.
E2800 Marco Deveronico All Flash or Hybrid system
Appro Xtreme-X Supercomputers
5th DOSAR Workshop Louisiana Tech University Sept. 27 – 28, 2007
GridTorrent Framework: A High-performance Data Transfer and Data Sharing Framework for Scientific Computing.
Presentation transcript:

Pathway to Petaflops A vendor contribution Philippe Trautmann Business Development Manager HPC & Grid Global Education, Government & Healthcare

Sun Proprietary/Confidential: Internal Use Only A Petaflop is : (10 15 ) floating point operations per second! Is there a limit to compute power requirements ? My own take is that the first PetaFlop system will be installed in 2007, latest 2008

Sun Proprietary/Confidential: Internal Use Only Sun HPC Vision To Create the Best Standards based Technologies, Products and Services Needed For High Performance Computing

Sun Proprietary/Confidential: Internal Use Only 1. Multicore CPUs are driving larger node sizes in terms of memory, I/O and power per node 2. Virtualization is increasing application footprint and therefore memory per node 3. InfiniBand and 10 GigE are offering low latency and increasingly cost-effective Fabrics 4. Power and cooling bills are sometimes more expensive than the system purchase cost 5. User/Researcher efficiency requires new programming tools and models Major Technology Trends Average compute node size is increasing in CPU cores, memory, I/O and power

Sun Proprietary/Confidential: Internal Use Only Top500 HPC Trends Clusters: > 72.8% of HPC Architecture Standards Based Interconnects: InfiniBand and Gigabit Ethernet > 80% of HPC Clusters Storage Capacity and Bandwidth Multi-TeraBytes to Multi-PetaBytes Parallel Filesystem Compute Nodes Multi-Core CPUs Moving toward Clusters of SMPs

Sun Proprietary/Confidential: Internal Use Only Tokyo Institute of Technology World’s Largest x64 Cluster Over 10,000 x64 processor cores, 20 TB main memory Over 1 PB Sun Storage with Lustre Parallel File System World’s Fastest Infiniband Network from Voltaire 6 weeks to deploy! Using N1SM and N1GE 38.16TFlops (Top500 #7), and expect further improvement at SC06

Sun Proprietary/Confidential: Internal Use Only 24Gbps (unidirection) Sun Fire x64 655nodes CPU Cores Memory: 21.4TB ClearSpeed Advance Board Storage Server B NEC iStorage S1800AT Physical Capacity 96TB RAID6 External Network Devices 64GB2.6GHz16B 639 nodes 2.4GHz CPU clock 32GB MEM/node A type ・・・・・・ x 42 Storage Server A Sun Thumper Physical Capacity 1PB Petabyte-class Storage Server 500GB 48disks 500GB 48disks InfiniBand Network Voltaire ISR 9288 ×8 100 Gigabit-class Network Equipment 500GB 48disks FileServ er CC Titech Top Level System Diagram

Sun Proprietary/Confidential: Internal Use Only Sun Fire X4500 “Storver” Selected by IN2P3 to run 400TB of data for the LHC-CERN High Performance Computing Data Server Compute ● 2 x Dual Core Opteron processors ● 16GB Memory Storage ● 48 Serial ATA disks ● Up to 24TB raw capacity I/O ● Very high throughput ● 2x PCI-X slots ● 4 GigE Availability ● Hot-swap power, fans, disks Management ● Same management as other Galaxy servers Solaris (TM) ZFS ● Ground-breaking file system performance Base Storage Configuration 48 x4500 data servers Operating Systems: Solaris/ZFS

Sun Proprietary/Confidential: Internal Use Only World's Storage Leader ● 67% of mainframe attached storage ● 37% of world's data ● Delivering NAS performance ● #1 WW Unix platform storage leader by PB delivered ● 8 out of 11 LHC Tier1 sites run StorageTek backup

Sun Proprietary/Confidential: Internal Use Only High Productivity Computing Systems Providing a new generation of economically viable high productivity computing systems for the national security and industrial user community ( )

Sun Proprietary/Confidential: Internal Use Only DARPA's “Ultrascale” Goals “May require up to 100,000 processor cores, but must be easier to use than current 1,000 processor systems.” 2+ PetaFlops on Linpack Top500 benchmark - Solves large linear system using Gaussian Elimination 6.5 PetaBytes/sec data streams bandwidth - Copy, add, scale large vectors locally 3.2 PetaBytes/sec bisection bandwidth - Transpose a large matrix 64,000 GigaUpdate /sec - Random updates to memory

Sun Proprietary/Confidential: Internal Use Only Sun and US Government Partnership Basic Computer Systems Research HPC Solution Center HPC Partner Programs HPC Software Year 3 Years 4 Years I: Industry Concept Study III: Product Development II: Research $ ~50M (July 2003) $ ~3M $ ~250M (End 2006) ??? 1 or 2 to Be Selected

Sun Proprietary/Confidential: Internal Use Only Sun's Qualification of Scientific Linux Scientific Linux 4.5 installed and ran successfully on Sun x64 systems > X2200, 2-socket, 4-core AMD x64 systems Scientific Linux 4.5 installed and ran successfully on Sun x64 Data Server at CERN > X4500, 2-socket, 4-core AMD x64 system with 48 SATA Disk Drives > Next test will be with Solaris and ZFS >

Sun Proprietary/Confidential: Internal Use Only Sun and DESY Collaboration dCache > Used World Wide in most of CERN Tier Sites > A distributed disk caching system as a front-end for Mass Storage System > Written in Java and until recently, only ran on Linux Sun worked with DESY to port dCache to Solaris x64 and ZFS > dCache ran on a x4500, half configured with 24 SATA Drives and 8GB of memory With Solaris and ZFS > No silent data corruption > Self-healing > Dynamic > High Performing

Sun Proprietary/Confidential: Internal Use Only Summary of Sun's Value to the HEP community Sun is Committed to CERN WW for Today and the Future We are demonstrating CERN Software stack on Sun Hardware and Solaris (Qualified Scientific Linux) Many of the HEP compute sites run Sun to manage Data through Sun StorageTek Technology Sun invests in collaborations with the community (Darpa, NSF, CERN LHC, European R&D projects, etc.) Collaborations allow creation of new tools and models that will allow researchers to run the PetaFlop systems

Thank you for your attention Philippe Trautmann Business Development Manager HPC & Grid Global Education, Government & Healthcare