[torqueusers] GPU Hardware

Lloyd Brown lloyd_brown at byu.edu
Thu Oct 21 10:10:00 MDT 2010


On 10/21/10 9:58 AM, Ken Nielson wrote:
> We would like to know what kind of GPU hardware is deployed in your data 
> centers so we can better understand what we need to do to provide GPU 
> support in TORQUE.
>
> If you currently have GPUs deployed would you please reply and let us 
> know the details of the hardware and any other information you think is 
> relevant.
>
> Thanks
>
> Ken Nielson
> Adaptive Computing
> _______________________________________________
> torqueusers mailing list
> torqueusers at supercluster.org
> http://www.supercluster.org/mailman/listinfo/torqueusers

We have two NVidia Tesla S1070s, attached via PCIe cables to two hosts,
meaning all 4 GPUs in each S1070, are available on a single host.  We
would love to have this separately schedulable from CPUs in the
Torque/Moab combination.  Basically, we'd love to have it detected as a
completely separate resource, in the same way that CPUs are detected and
managed separately from RAM, or Disk, etc.  At the moment we're
kinda/sorta faking it using GRES, but it has a few problems.

We're going to be testing a Fermi card as well, but it's not deployed yet.


-- 


Lloyd Brown
Systems Administrator
Fulton Supercomputing Lab
Brigham Young University
http://marylou.byu.edu




More information about the torqueusers mailing list