Logo

Slurm

Connect training jobs, allocations, users, and cluster scheduling metadata.

Built by

Resource

Overview

Slurm is commonly used in research and training clusters. Matcha uses Slurm job metadata to connect GPU power telemetry with training jobs, users, queues, and allocations, making it easier to understand energy by run or research workload.

Configuration Steps

  1. Connect Matcha to Slurm job accounting data.

  2. Map job IDs to GPU allocations.

  3. Import user, queue, and runtime metadata.

  4. Align job windows with GPU telemetry.

  5. Review energy by job and allocation.