Uploaded image for project: 'Performance and Scale for AI Platforms'
  1. Performance and Scale for AI Platforms
  2. PSAP-1441

Add Tensor core usage to MIG partitioning experiments

XMLWordPrintable

    • Icon: Story Story
    • Resolution: Unresolved
    • Icon: Normal Normal
    • RHOAI_2.11.0
    • None
    • None
    • None
    • Inference, RHELAI
    • False
    • False
    • None
    • 3
    • PSAP - General-7, PSAP - General-8, PSAP - General-9, PSAP - General-10, PSAP - General-11, PSAP - General-12

      Summary: Add Tensor core utilization to MIG partitioning experiments

      Description: Currently, the MIG (Multi-Instance GPU) partitioning experiments do not include Tensor core usage analysis. This card aims to incorporate Tensor core utilization metrics into our existing MIG partitioning experiments to evaluate performance improvements and feasibility.

      Tasks:

      1. Research Tensor core utilization capabilities within current MIG setups.

      Acceptance Criteria:

      • Tensor core utilization successfully integrated into MIG partitioning experiments.
      • Performance metrics collected and analyzed for comparison with baseline.
      • Documentation and report on findings completed and shared with stakeholders.
      • Validation of results through peer review and discussion.

              ccamacho@redhat.com Carlos Camacho
              ccamacho@redhat.com Carlos Camacho
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

                Created:
                Updated: