RE: Workstation Computers for FLUKA

From: Paul Colin Gloster <Colin_Paul_Gloster_at_ACM.org>
Date: Mon, 29 Nov 2010 17:52:54 +0000 (UTC)

Nicholas Bolibruch <nboli_at_Acanac.net> sent on November 26th, 2010:
|----------------------------------------------------------------------|
|"I don't know if you've looked at the performance of GPU cards lately,|
|[..] |
|Following Moore's Law, using a similar configuration, in 2 years that |
|will be 4 TFLOPS, and in 4 years it will be 8 TFLOPS. |
| |
|[..]" |
|----------------------------------------------------------------------|

Hi,

If I may nitpick without lessening the point that G.P.U.s may be worth
considering... on April 13th, 2010 I attended a presentation in which
it was claimed that for linear algebra at least, the speeds of G.P.U.s
increase to 2 times the speeds every year (since 2008) in contrast to
"multi"core machines' speeds increasing to 1.5 times the speeds every
year.

Regards,
Paul Colin
Received on Mon Nov 29 2010 - 23:27:07 CET

This archive was generated by hypermail 2.2.0 : Mon Nov 29 2010 - 23:27:08 CET