DGX-A100 blog post #1
Conversation
kpouget
force-pushed
the
dgx
branch
2 times, most recently
from
40fa585 to
d9cdb97
Compare
| help of NFD labels, the GPU Operator will automatically detect the DGX | ||
| A100 node and install (via containers) the necessary drivers, services | ||
| and CUDA libraries to run GPU workload. | ||
| Do not forget to create the default `ClusterPolicy` custom resource |
There was a problem hiding this comment.
It would be good to explain why you need to create the default ClusterPolicy instead of just staying don't forget to do it. Also there is a '.' missing at the end of the sentence.
There was a problem hiding this comment.
hum, now I'm wondering if it makes sense to keep it ... There's a link to the documentation, maybe it doesn't make sense to stress a particular point
There was a problem hiding this comment.
I think if it's a common mistake I would keep it. Either way I think it's good to have one sentence explaining why you are creating the default ClusterPolicy. It makes the blog more complete and understandable.
There was a problem hiding this comment.
changed to:
Do not forget the last step, the creation of the
ClusterPolicycustom resource! This object defines the configuration of the GPU stack, and without it, nothing will be deployed on the nodes!
Co-authored-by: David Gray <40244437+dagrayvid@users.noreply.github.com>
Request from NVIDIA Product Management to remove the dash
Good to place the product name in the conclusion. Fix DGX A100 name.
No description provided.