[CP2K:5122] Re: how to improve CP2K scaling for small systems

Alin Marin Elena alinm... at gmail.com
Mon Mar 31 09:28:27 UTC 2014


Thanks I will give it a spin these days.

I do not have access to the memory banks.

Alin
On 31 Mar 2014 10:08, "labicia" <lab... at yahoo.it> wrote:

> Please, find attached the requested file.
>
> If you have the chance, it will be good to see the scaling per core.
>
> But even more interesting it would be to see the scaling per memory
> channel, keeping the number of cores fixed to the maximum available. This
> is a bit more tricky, though, and requires physical access to the computer
> ram.
> PROCEDURE
> 1) Remove all RAM banks apart from the ones belonging to channel 1. At
> this point, the amount of RAM available (in channel 1) becomes the limiting
> factor to launch the ZIF8 simulation. If needed, one has to decrease the
> CUTOFF until the simulation can be performed, and store its timing.
> 2) Populate also the RAM second channel and repeat the simulation.
> 3) Do the same for channel 3 and then for channel 4.
>
> It will be interesting to see how the timing evolves.
>
> Thanks a lot,
> Marco and Andrea
>
> --
> You received this message because you are subscribed to the Google Groups
> "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to cp2k+uns... at googlegroups.com.
> To post to this group, send email to cp... at googlegroups.com.
> Visit this group at http://groups.google.com/group/cp2k.
> For more options, visit https://groups.google.com/d/optout.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20140331/350a3e16/attachment.htm>


More information about the CP2K-user mailing list