Talk:RIKEN MDGRAPE-3

Tip: Anchors are case-sensitive in most browsers.

This article links to one or more target anchors that no longer exist.

[[Xeon#Dual-Core Xeon .2865 nm NetBurst.29|"Dempsey"]]

Please help fix the broken anchors. You can remove this template after fixing the problems. | Reporting errors

Expansion

For people interested in expanding the article... http://www.newsfactor.com/news/Japan-Bests-IBM-in-Supercomputer-Stakes/story.xhtml?story_id=1220059R0ADY --Oblivious 14:05, 30 July 2006 (UTC)[reply]

Does anybody know why exactly MDGRAPE-3 is incapable of running LINPACK? What kind of software is this system running on? dreddnott 22:06, 10 August 2006 (UTC)[reply]

Well it would be able to run, but couldn't use the power of the special CPUs, so it would be extremly slow. This is because the special CPUs can only compute particle interactions. LINPACK is linear algebra.--Rolandschulz 17:55, 7 February 2007 (UTC)[reply]

Expansion 2

What are the physical dimensions? -- Beland 17:40, 3 June 2007 (UTC)[reply]

MDGRAPE instead of RIKEN_MDGRAPE-3?

There's a predecessor: MDGRAPE-2 and the Molecular Dynamics Machine (MDM) It seems that MDGRAPE-3 is also a processor (MDGRAPE-3 chip) and a PCIX-Expansion (MD3-PCIX) card and not only a system! (see http://mdgrape.gsc.riken.jp/modules/tinyd0/index.php Achievements)

So why not have an article about the MDGRAPE Project (including the predecessor MDGRAPE-2)? —Preceding unsigned comment added by 141.58.5.85 (talk) 10:56, 14 February 2008 (UTC)[reply]