Jump to: navigation, search

Difference between revisions of "ScientificWGGPUs"

(Activities)
(Activities)
Line 11: Line 11:
  
 
* Initial setup of GPU environment with passthrough configured per PCI device. (Blog: http://www.vscaler.com/gpu-passthrough)
 
* Initial setup of GPU environment with passthrough configured per PCI device. (Blog: http://www.vscaler.com/gpu-passthrough)
* M3 cluster production GPU passthrough deployment at Monash University.
+
* M3 cluster production GPU passthrough deployment at Monash University
 +
* OVH gave a nice talk at the Sydney summit - https://www.youtube.com/watch?v=1tdvz3ejokM
 +
* For some GPUs, the hypervisor signature needs to be hidden
 +
** By image,
 +
** By flavor, patch - https://review.openstack.org/#/c/555861/
 
* Performance evaluation of the GPUs within a virtualised environment in comparison to a bare metal environment.
 
* Performance evaluation of the GPUs within a virtualised environment in comparison to a bare metal environment.
 
* Upstream guide docs on how to build a GPU-enabled OpenStack cloud
 
* Upstream guide docs on how to build a GPU-enabled OpenStack cloud
 
** initial draft notes underway at https://etherpad.openstack.org/p/scientific-wg-gpu-config-guide
 
** initial draft notes underway at https://etherpad.openstack.org/p/scientific-wg-gpu-config-guide
 
** collating details of known working and not-working GPU-hardware-hypervisor combinations at https://etherpad.openstack.org/p/GPU-passthrough-model-success-failure
 
** collating details of known working and not-working GPU-hardware-hypervisor combinations at https://etherpad.openstack.org/p/GPU-passthrough-model-success-failure

Revision as of 06:11, 20 April 2018

Context

Using GPUs as accelerators for application workloads.

People

  • David Power (powerd)
  • Blair Bethwaite (b1airo)

Resources

Activities