Re: [AMBER] Horrific pmemd.cuda performance on an 8-GPU system

From: michael <michael.bane.manchester.ac.uk>
Date: Wed, 11 Aug 2010 22:48:36 +0100

On Wed, 2010-08-11 at 12:15 -0700, Sasha Buzko wrote:
> Hi guys,
> We just got an 8-GPU system that uses onboard PLX switches that expand
> four x16 PCIe slots on a Tyan board to eght. I ran a small production
> simulation test, and the results are horrifying, to say the least.
> If you could help narrow down possible sources of the problem, it would
> be great.

A shot in the dark here, but did you check the reference system
completed the simulation (and both systems gave same results)? I ask
only since I see that the ref sys has
 Run time limit on kernels=YES
but this isn't true for the new sys

M.


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Aug 11 2010 - 15:00:05 PDT
Custom Search