NVIDIA Management Library (NVML) is a C-based API for monitoring and managing NVIDIA GPU devices. NVML go bindings are taken from nvidia-docker 1.0 with some improvements and additions. NVML headers are also added to the package to make it easy to use and build.
Three samples are included to demonstrate how to use the NVML API.
GPU metrics exporter for Prometheus leveraging NVIDIA Data Center GPU Manager (DCGM) is a simple shell script that starts nv-hostengine, reads GPU metrics every 1 second and converts it to a standard Prometheus format.
Find the installation and run instructions here.
A signed copy of the Contributor License Agreement needs to be provided to [email protected] before any change can be accepted.
- Please let us know by filing a new issue
- You can contribute by opening a pull request