Most of the computing power these comes from accelerators (GPUs and Xeon Phis). For using GPUs from a docker container, Regan suggests a solution using Ubuntu (http://stackoverflow.com/a/25367554/2462245). Has anyone has tried a similar setup to use Xeon Phis from a container? Does anyone think/know that running codes on accelerators from containers will hit performance?
On Thu, May 21, 2015 at 9:30 AM, Kilian Cavalotti < [email protected]> wrote: > Hi Fabricio, > > On Wed, May 20, 2015 at 6:18 PM, Fabricio Cannini <[email protected]> > wrote: > > What are your opinion(s) about the subject? Good idea or overkill ? > > http://www.theplatform.net/2015/03/23/hpc-schedulers-snap-to-docker/ > > We're precisely having that discussion on the slurm-dev mailing-list: > https://groups.google.com/forum/#!topic/slurm-devel/oPmz5em5tAA > > Cheers, > -- > Kilian > _______________________________________________ > Beowulf mailing list, [email protected] sponsored by Penguin Computing > To change your subscription (digest mode or unsubscribe) visit > http://www.beowulf.org/mailman/listinfo/beowulf >
_______________________________________________ Beowulf mailing list, [email protected] sponsored by Penguin Computing To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
