1 / 9

SOS7 What will Cray do for Supercomputing in this Decade?

SOS7 What will Cray do for Supercomputing in this Decade?. Asaph Zemach Cray Inc. I am not Burton. Sorry. Where is Burton?. Maybe?. More likely…. An Apology. Cray MTA-2 Accepted by NRL Sept ‘02 UMA Shared Memory Latency Tolerant: 128 contexts in processor. Red Storm for Sandia

braith
Download Presentation

SOS7 What will Cray do for Supercomputing in this Decade?

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. SOS7What will Cray do for Supercomputing in this Decade? Asaph Zemach Cray Inc

  2. I am not Burton. Sorry. Where is Burton? Maybe? More likely… An Apology

  3. Cray MTA-2 Accepted by NRL Sept ‘02 UMA Shared Memory Latency Tolerant: 128 contexts in processor. Red Storm for Sandia Contract signed Oct ‘02 10,000 AMD X86-64 High Speed Network Cray X1 FCS Dec 31, 2002 Scalable vector MSPs NUMA Shared Memory What Have You Done For Me Lately?

  4. Cray Products: The Near Future 2003 End of ‘04 2005+ X112.8 GF35GB/s/p mem BW76GB/s/p cache BW X1eTechnology UpgradeFaster clockDenser PackageMix&Match with X1 X2(Blackwidow)BiggerFasterCheaper Synergy? I/O? 2003 End of ‘04 2005 Red Storm(Development) Red Storm(Install)Catamount LWKLinux serviceAMD 2GHz X86-64 Red StormProduct (?)Linux ServiceCompute OS?

  5. Shared Memory Locales UMA, NUMA Heavy Weight Processors Multi threading, Vectors, Streams PIM (LWP) Cray Products: Not So Near Future 2006(?) 2008(?) 2010 Cascade X2eBIGGERFASTERCHEAPER X2fBIGGER!!FASTER!!CHEAPER!!

  6. Cascade Locale Heavy Weight Proc Vector MT Streams SW Controlled Data Cache Multithreaded PIM DRAM Multithreaded PIM DRAM Locale Interconnect Multithreaded PIM DRAM Multithreaded PIM DRAM Multithreaded PIM DRAM Multithreaded PIM DRAM Router To other Locales

  7. Cascade: Lazy Localization Memory • Initially all data is considered generic – equally far from everywhere. • To improve performance stage generic data near HWP that manipulates it. • To improve performance even more, partition data between PIMS. • All data always universally accessible but performance varies. Generic Data Highly Localized Data Somewhat Localized Data HWP

  8. Compiler controlled cache Compartmentalized OS-es Introspection using PIM Relative Debugging Abstract locales: virtualize locality management What needs to be near what What can/should be distributed (& how) Cascade: Software Investigations

  9. Burton Smith David Callahan Steve Scott Cray Thomas Sterling Larry Bergman Hans Zima JPL, CalTech Jay Brockman Peter Kogge Notre Dame Bill Daly Stanford Cascade People

More Related