AMD Instinct Data Center GPU Documentation

Targeted for system administrators and power users, this site provides comprehensive technical documentation, guides, and best practices for deploying and managing AMD Instinct Data Center GPUs. Content for GPU software development and applications are located at ROCm documentation.

Beta status

This website is in beta status. The information here is accurate to our best efforts. The organization, layout and structure of this site will change rapidly. Feedback is appreciated via the ROCm discussions forum.

System deployment and configuration

Cluster Networking

Optimize the network for Instinct GPU applications

System Acceptance Test Guide

Detailed instructions to test the proper functioning and optimal performance of server systems equipped with AMD Instinct MI300X GPU accelerators.

Kubernetes

AMD Kubernetes Device Plugin

Kubernetes (k8s) device plugin to enable registration of AMD GPU to a container cluster

AMD GPU Operator

The AMD GPU Operator simplifies the deployment and management of AMD Instinct GPU accelerators within Kubernetes clusters.

Telemetry

AMD Device Metrics Exporter

AMD Device Metrics Exporter enables Prometheus-format metrics collection for AMD GPUs in HPC and AI environments.