Uploaded image for project: 'Red Hat OpenStack Services on OpenShift'
  1. Red Hat OpenStack Services on OpenShift
  2. OSPRH-12310

Document support matrix for gpu vendor vs gpu model vs passthrough type

XMLWordPrintable

    • Icon: Story Story
    • Resolution: Done
    • Icon: Undefined Undefined
    • None
    • None
    • VAF
    • None
    • 3. RHEL AI Automation
    • 1

      As an Engineer and Product manager:
      I want to understand which of the gpu models from different vendors have been tested, should be supported, and in what configuration.

      So that:
      We can provide information to customers on current support level as well information to the team on work already done and know-how.

      Definition of Done:

      • Define a matrix consisting of:

      – GPU vendor
      – GPU model
      – passthrough type -> vGPU vs timeslicing vs PCI passthrough
      – version of drivers used

      • Version of components used (OS, Libvirt, firmware, etc...)
        – other...
      • Add information on how/where these accelerators have been tested
      • Add information on customers using these accelerators

      Dependencies:

      • Information from PMs on customers using accelerators in RHOSO

      Deliverables:

      • Confluence that will be updated with test of new accelerators

      Acceptance Criteria:

      • Automated/Tested/Supported (based on card X)/Not supported statement for each combination

       

              erpeters@redhat.com Erin Peterson
              lsvaty@redhat.com Lukas Svaty
              rhos-workloads-vaf
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

                Created:
                Updated:
                Resolved: