[torqueusers] gpgpu

Brock Palen brockp at umich.edu
Wed Feb 24 07:53:30 MST 2010


On our system due to there being only 4 gpus, per node, and a node has  
8 cores, we used Moab's node locked resource, and treat it like a  
software token.  Each node has 4 'graphics' resources, they can  
request.  When all 4 are given out regular cpu jobs can run on the  
other 4 cores.  We also set the node priority to something lower than  
the normal nodes, so regular cpu jobs don't eat up GPU nodes until  
there are nothing else left.

Brock Palen
www.umich.edu/~brockp
Center for Advanced Computing
brockp at umich.edu
(734)936-1985



On Feb 24, 2010, at 2:07 AM, Dr. Stephan Raub wrote:

> Hi,
>
> I have set up a cluster once, where I defined arch="cuda" in the mom- 
> config
> of everynode, that has build in a TESLA card. Then I created a queue  
> named
> "cuda" which accepts only jobs with arch="cuda" in the resource
> requirements. Maui takes care, that only one cuda job is scheduled  
> on such a
> node. I asked in the Mauiusers-Mailinglist (10.12.2009) how to do  
> this best
> and got the answer to set
>
> QOSCFG[monopol]          QFLAGS=DEDICATED
> CLASSCFG[cuda]           QLIST=monopol QDEF=monopol
>
> Does that match your needs?
>
> Stephan
> --
> ---------------------------------------------------------
> | | Dr. rer. nat. Stephan Raub
> | | Dipl. Chem.
> | | Lehrstuhl für IT-Management / ZIM
> | | Heinrich-Heine-Universität Düsseldorf Universitätsstr. 1 /
> | | 25.41.O2.25-2
> | | 40225 Düsseldorf / Germany
> | |
> | | Tel: +49-211-811-3911
> ---------------------------------------------------------
>
> Wichtiger Hinweis: Diese E-Mail kann Betriebs- oder  
> Geschäftsgeheimnisse,
> bzw.
> sonstige vertrauliche Informationen enthalten. Sollten Sie diese E- 
> Mail
> irrtümlich erhalten haben, ist Ihnen eine Kenntnisnahme des Inhalts,  
> eine
> Vervielfältigung oder Weitergabe der E-Mail ausdrücklich untersagt.  
> Bitte
> benachrichtigen Sie uns und vernichten Sie die empfangene E-Mail.  
> Vielen
> Dank.
>
> Important Note: This e-mail may contain trade secrets or privileged,
> undisclosed or otherwise confidential information. If you have  
> received this
> e-mail in error, you are hereby notified that any review, copying or
> distribution of it is strictly prohibited. Please inform us  
> immediately and
> destroy the original transmittal. Thank you for your cooperation.
>
>
>> -----Ursprüngliche Nachricht-----
>> Von: torqueusers-bounces at supercluster.org [mailto:torqueusers-
>> bounces at supercluster.org] Im Auftrag von Timothy Lindsay
>> Gesendet: Dienstag, 23. Februar 2010 22:07
>> An: torqueusers at supercluster.org
>> Betreff: [torqueusers] gpgpu
>>
>> I am researching the best way to schedule GPGPU jobs.  What type of
>> support, if any, does Torque have for GPGPU?
>>
>> ----------------------------------------------------------------
>> This message was sent using IMP, the Internet Messaging Program.
>>
>>
>> _______________________________________________
>> torqueusers mailing list
>> torqueusers at supercluster.org
>> http://www.supercluster.org/mailman/listinfo/torqueusers
>
>
> _______________________________________________
> torqueusers mailing list
> torqueusers at supercluster.org
> http://www.supercluster.org/mailman/listinfo/torqueusers
>
>



More information about the torqueusers mailing list