Re: [AMBER] Analysis of pmemd performance for exascale

From: David A Case via AMBER <amber.ambermd.org>
Date: Thu, 5 Oct 2023 08:41:51 -0600

On Thu, Oct 05, 2023, David Case wrote:
>>
>>As part of an Exascale project I’ve been asked to gather information
>>on biomolecular simulation code performance in terms of numerical
>>methods.

Just a quick addition here: as it currently exists, Amber is not really
suited for exaxcale computation. It has been written and optimized for
execution on one (or a very small number) of inexpensive GPU cards. Other
simulation packages are much more targetted toward more powerful (and more
expensive) HPC hardware.

Some applications, such as replica exchange, can make use of many GPUs in
the context of loosely parallel execution.

....dac


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Oct 05 2023 - 08:00:03 PDT
Custom Search