NVIDIA Requirements for AI Clouds

Appendix

View as Markdown

Appendix

This section contains links to reference documents and implementation guides to provide additional details if NCPs need them.

Test Legend

As a part of the AI Cloud-Ready Initiative, a validation suite is provided on GitHub, containing a set of tests that can be run to evaluate an environment against the provided requirements and specifications.

The Test Details column in the following sections contains four options:

### - Github Test reference
TBD - item on the github roadmap
add - item to be added into the roadmap
INFO - a requirement which is either not testable or currently won’t be tested from the test repo.

Implementation Guidance

Reference documents provide additional information on implementing some of the above requirements.

  1. Network Topology Discovery: https://github.com/NVIDIA/topograph
  2. Exemplar Cloud Website: https://www.nvidia.com/en-us/data-center/ai-cloud-performance/
  3. Kubernetes Security guidance : https://github.com/kubernetes/committee-security-response

Other Feature Considerations (Not Required)

  1. Disk Cloning: Disk cloning capability (for network-attached block devices). It should be possible to clone a disk even on a running instance.