nvidia_gpu_exporter

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

---

Warning

Maintenance Status: I get that it can be frustrating not to hear back about the stuff you've brought up or the changes you've suggested. But honestly, for over a year now, I've hardly had any time to keep up with my personal open-source projects, including this one. I am still committed to keep this tool working and slowly move it forward, but please bear with me if I can't tackle your fixes or check out your code for a while. Thanks for your understanding.

---

Introduction

There are many Nvidia GPU exporters out there however they have problems such as not being maintained, not providing pre-built binaries, having a dependency to Linux and/or Docker, targeting enterprise setups (DCGM) and so on.

This is a simple exporter that uses nvidia-smi(.exe) binary to collect, parse and export metrics. This makes it possible to run it on Windows and get GPU metrics while gaming - no Docker or Linux required.

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

If you are a gamer who's into monitoring, you are in for a treat.

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 the monitored machine: can be configured to execute nvidia-smi command remotely
• No need for a Docker or Kubernetes environment
• Auto-discovery of the metric fields nvidia-smi can expose (future-compatible)
• Comes with its own Grafana dashboard

Visualization

You can use the official Grafana dashboard to see your GPU metrics in a nicely visualized way.

Here's how it looks like:

Installation

See INSTALL.md for details.

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:

./testdata/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 testdata/captures/README.md.

Star History