Support process for GPU failure in HGX system under NVIDIA AI Enterprise Business Standard

Hello,

I have an HGX system with 8 GPUs, and I purchased NVIDIA AI Enterprise licenses for all of them. Suppose one of the GPUs fails — under the Business Standard Support plan, what kind of support can I expect?

For example:

  • Is the failed GPU replaced via RMA (I ship it back and receive a new one)?

  • Does an NVIDIA representative come onsite to replace it?

  • Or is there another procedure for handling hardware failures under this support plan?

Thank you in advance for clarifying how GPU-level hardware issues are handled in this situation.

Hi there @ugurbas!

Any of those questions will be answered by our dedicated Enterprise support team.

Please open a case to get that information.

Thank you!