gpu-telemetry

Enrichment pending

GPU Observability with workload attribution. One OTLP agent per node ties hardware metrics (NVIDIA, AMD, Intel Gaudi) to the K8s pod or Slurm job burning the GPU.

GitHub Website

GraphCanon updated today · GitHub synced today

56 stars6 forksLast push 1w Python MIT

gpu-telemetry alternatives Trust report

Verify the decision

Maintenance and security

Full trust report

Maintenance: Active (8d since push); As of today
Provenance: Not a fork · Organization account; As of today
Security (OSV): No lockfile; As of today

Public GitHub metadata and optional OSV scans. Signals, not a guarantee. Trust methodology.

Install

pip install gpu-telemetry

PyPI

Similar tools

Same-category neighbours. No typed graph edges are catalogued for this tool yet.

ECC

The agent harness performance optimization system for AI agents

228k★

hermes-agent

The agent that grows with you

213k★

AutoGPT

AutoGPT is the vision of accessible AI for everyone, to use and to build on.

185k★

Evidence and technical details

Sourced facts, taxonomy, compatibility claims, README excerpt, and machine-readable endpoints.

Overview

GPU Observability with workload attribution. One OTLP agent per node ties hardware metrics (NVIDIA, AMD, Intel Gaudi) to the K8s pod or Slurm job burning the GPU.

Capability facts

CLI: CLI entrypoint
Source: pyproject.toml:[project.scripts] · Jul 15, 2026
Languages: python
Source: github.language+pyproject.toml · Jul 15, 2026

Compatibility

Sourced claims from the README excerpt - not unsourced marketing copy.

Python runtimePython

Source: README excerpt (regex_v1, Jul 15, 2026)

pip install l9gpu

Source link

README

Quick Start — Bare Metal / systemd

pip install l9gpu
export OTEL_EXPORTER_OTLP_ENDPOINT=<your-otlp-endpoint>
export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Bearer <your-token>"

l9gpu nvml_monitor  --sink otel --cluster my-cluster  # NVIDIA
l9gpu amd_monitor   --sink otel --cluster my-cluster  # AMD
l9gpu gaudi_monitor --sink otel --cluster my-cluster  # Intel Gaudi

Sanity-check without OTLP: l9gpu nvml_monitor --sink stdout --once.

systemd unit files: systemd/.

Hardware support

NVIDIA A100, H100 / H200, B200 / GB200, T4, A10, L4 (NVML + DCGM) · AMD MI300X, MI325X (amdsmi) · Intel Gaudi 2, Gaudi 3 (hl-smi).

Full metric catalog with units and attributes: docs/METRICS.md.

License

MIT for l9gpu, k8shelper, k8sprocessor. Apache-2.0 for slurmprocessor, shelper. Each subdirectory carries its own LICENSE where it differs from the repo root.

For agents

This page has a .md twin and JSON over the API.

/tools/last9-gpu-telemetry.md GET /api/graphcanon/tools/last9-gpu-telemetry GET /api/graphcanon/graph?tool=last9-gpu-telemetry