[VGPU] Multiple VGPUs on one VM

Jahson Babel jahson.babel at cc.in2p3.fr
Wed Jan 19 08:54:18 UTC 2022


Hello folks,

I've seen a thread on vGPUs on this mailing list lately and I was 
wondering if multiple vGPUs are supported on one VM on OpenStack.

 From the OpenStack's documentation this seems pretty clear that it's 
not. ( https://docs.openstack.org/nova/latest/admin/virtual-gpu.html)
But on Nvidia's side multiple vGPUs  should be able to be possible on 
specific Linux KVM version. 
(https://docs.nvidia.com/grid/latest/grid-vgpu-release-notes-red-hat-el-kvm/index.html#multiple-vgpu-support)
Only true for RHEL systems maybe.

So which one of those pieces of information should I follow ? Presumably 
the OpenStack's side is the correct one but I would like to be sure.
And does anyone has been able to attach multiples vgpus on one VM ?
Another side question that I have in mind despite the fact that I can't 
see how it could possibly work. Is it possible to over-commit the vGPUs 
resources ? Via the placement API ?

Thanks in advance.

Kindly,

Jahson Babel


-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4261 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://lists.openstack.org/pipermail/openstack-discuss/attachments/20220119/e3f45045/attachment.bin>


More information about the openstack-discuss mailing list