NASA announces world's fastest supercomputer
NASA announces world's fastest supercomputer
Posted Oct 26, 2004 22:44 UTC (Tue) by JoeBuck (subscriber, #2330)In reply to: NASA announces world's fastest supercomputer by vondo
Parent article: NASA announces world's fastest supercomputer
Actually, I believe that Lawrence Livermore Lab has ordered a version of the IBM BlueGene/L machine that should pass the SGI/NASA machine when it is completed. So yes, someone has ordered one.
Posted Oct 26, 2004 23:07 UTC (Tue)
by emkey (guest, #144)
[Link]
http://www.serverwatch.com/news/article.php/3416891
And the IBM benchmark was apparently on 1/10th of the total system to be delivered.
Posted Oct 26, 2004 23:53 UTC (Tue)
by dcg (subscriber, #9198)
[Link] (3 responses)
Posted Oct 27, 2004 0:50 UTC (Wed)
by emkey (guest, #144)
[Link] (2 responses)
Still, an impressive accomplishment.
Posted Oct 27, 2004 5:16 UTC (Wed)
by gurulabs (subscriber, #10753)
[Link]
Posted Oct 27, 2004 7:36 UTC (Wed)
by pointwood (guest, #2814)
[Link]
:)
Thats my understanding as well based on the following...NASA announces world's fastest supercomputer
According to http://www.sgi.com/company_info/newsroom/press_releases/2... , they only used 16 of the 20 Altix boxes, or 3072 less processors. I guess they could get higher figures.NASA announces world's fastest supercomputer
I'm curious as to why they only used 16. People generally pull out all the stops to get the highest possible numbers. The fact that 16 is a power of 2 makes me wonder if there is some sort of hardware/software issue that prevents them from using the full system.NASA announces world's fastest supercomputer
Why not announce as soon as you've broken the record vs waiting another X days before the whole thing is installed.NASA announces world's fastest supercomputer
This should answer your question: http://news.com.com/Two+records+in+one+day+for+SGI+superc...NASA announces world's fastest supercomputer