Show HN: L9gpu – GPU telemetry that ties each GPU to the K8s pod or Slurm job
Category: observability
Tags: gpu-monitoring, kubernetes, slurm, opentelemetry, observability
Score: 8.0/10 (Innovation: 7, Technical: 8, Documentation: 9, Utility: 8)
l9gpu provides GPU telemetry with workload attribution, tying each GPU to the Kubernetes pod or Slurm job that is using it. It is interesting because it solves a common pain point in large-scale GPU clusters by bridging hardware metrics and higher-level orchestration, using OpenTelemetry for vendor and backend agnosticism.
Target audience: devops, platform engineers, data engineers, ML engineers
Repository: https://github.com/last9/gpu-telemetry · Python · MIT · 47 stars
View on Hacker News