nvidia_gpu_exporter

Nvidia GPU exporter for prometheus, using nvidia-smi binary to gather metrics.

---

Warning

Heads up: this is a side project I maintain in my spare time. I might take a long time to look at issues or PRs, or not get to them at all. Sorry in advance, and thanks for understanding.

---

Introduction

This is a simple exporter that uses the nvidia-smi(.exe) binary to collect, parse and export metrics. Since it only needs nvidia-smi, it also works on Windows - no Docker or Linux required.

It can also skip nvidia-smi and read the metrics straight from the NVIDIA Management Library (NVML). This mode is experimental and exposes some things nvidia-smi cannot provide, like per-MIG-instance metrics and XID error counters; see CONFIGURE.md.

This project is based on a0s/nvidia-smi-exporter. However, this one is written in Go to produce a single, static binary.

Use cases

• Consumer and prosumer GPUs (GeForce/RTX), where the datacenter tooling exposes little and nvidia-smi is often the only uniform source of utilization, memory, power and temperature
• Small Kubernetes clusters, edge boxes and homelabs that want GPU metrics without installing the NVIDIA GPU Operator stack
• Virtualized or restricted setups (vGPU guests, MIG slices, locked-down containers) where the deeper GPU counters are not exposed but nvidia-smi still answers
• Mixed fleets of old and new cards that need one exporter that behaves the same everywhere
• Gaming rigs, for watching your GPU stats on a dashboard while you play

If you run datacenter cards on Kubernetes with the GPU Operator already installed, DCGM-exporter is probably the better fit; this exporter aims at the cases above.

Highlights

• Will work on any system that has nvidia-smi(.exe)? binary - Windows, Linux, MacOS... No C bindings required
• Doesn't even need to run on the monitored machine: can be configured to execute nvidia-smi command remotely
• Auto-discovery of the metric fields nvidia-smi can expose (future-compatible)
• Optional per-process GPU metrics: see which process uses how much GPU memory
• Optional background collection: run nvidia-smi on a timer instead of on every scrape
• Comes with its own Grafana dashboards: a per-GPU detail one and a multi-GPU overview

Try it without a GPU

Demo mode serves realistic synthetic metrics, including the NVML-only families, with no GPU, driver or even Linux required:

nvidia_gpu_exporter --collect.backend demo

By default it simulates two H200 GPUs with fluctuating values, a MIG topology and an XID error history. The simulated setup is configurable; see CONFIGURE.md.

Visualization

There are two official Grafana dashboards, and they link to each other in Grafana:

• Nvidia GPU Metrics (ID 14574), the per-GPU detail view.
• Nvidia GPU Overview (ID 25547), which compares all GPUs of a node side by side and drills down into the detail dashboard.

Import either by ID in Grafana (Dashboards - New - Import), or enable grafanaDashboard in the Helm chart to get both provisioned automatically. The JSON is also in this repository under docs/grafana.

Here's how they look:

Installation

You can install it from plain binaries, deb/rpm packages, winget, Docker images or the Helm chart. See INSTALL.md for details.

Verifying releases

Release artifacts are signed so you can check they came from this project's release pipeline:

• The checksums.txt file attached to each release is signed with GPG (checksums.txt.asc), which covers every binary, archive and package.
• The container images and the Helm chart are signed keyless with cosign, tied to the release workflow's identity.

See INSTALL.md for the exact verification commands, and the chart README for the chart.

Configuration

See CONFIGURE.md for details.

Metrics

See METRICS.md for details.

Contributing

See CONTRIBUTING.md for details.

Help wanted: contribute a GPU capture

The exporter parses nvidia-smi output, which differs across GPU models, driver versions and operating systems. If you have hardware that isn't covered yet (datacenter cards, MIG, multi-GPU, Windows/WSL2, brand-new drivers...), you can help a lot by capturing your nvidia-smi output with one command:

./internal/captures/collect.sh          # add --load for an under-load sample too

It needs only nvidia-smi, bash, and the standard core utilities (awk, sed, ...), runs read-only, and masks identifiers (GPU UUID, serial, hostname) by default. It writes one .txt file: commit it and open a PR, or attach it to an issue. See internal/captures/README.md.

Star History