AMD Instinct™ · Data Center GPU

Systems & Infrastructure Documentation

Everything you need to deploy, validate, and operate AMD Instinct™ Data Center GPUs at scale — drivers, orchestration, cluster management, and acceptance testing for HPC and AI. For API and software-stack reference, see the ROCm documentation.

Install the GPU driver Customer Acceptance Guide

Start here

System Administrators

Deploy and run AMD Instinct GPUs on bare metal, in containers, and across clusters. These guides are the most frequently updated content on this site.

Bare metal

Instinct GPU Driver Install and configure the GPU, including logging and error codes. GPU Partitioning Split compute units and memory to partition a single GPU. AMD SMI Unified user-space tool to manage and monitor GPUs and drivers. ROCm Validation Suite System validation and hardware diagnostics. Customer Acceptance Guide Configure, validate, benchmark, and baseline Instinct GPUs. Cluster Validation Suite Test scripts that validate AMD AI clusters end to end.

Containers & orchestration

GPU Operator Deploy and manage Instinct GPUs in Kubernetes clusters. Network Operator Simplify AMD AINICs in Kubernetes environments. Device Plugin Register AMD GPUs with a Kubernetes container cluster. Device Metrics Exporter Prometheus-format GPU metrics for HPC and AI environments. AMD Container Toolkit Integrate Instinct GPUs with Docker and container runtimes. Spur AI-native job scheduler, drop-in compatible with Slurm, with GPU-first scheduling and Raft-based state.

Cluster, cloud & virtualization

Enterprise AI Tools to manage enterprise AI infrastructure at scale. Omnistat Profile GPU resource utilization across the cluster. Cluster Networking Optimize the network for Instinct GPU applications. Instinct on Azure Get started with AMD Instinct GPUs on Microsoft Azure. Virtualization Driver Explore the virtualization driver for Instinct GPUs. AMD SMI for Virtualization Manage and monitor virtualization-enabled AMD GPUs.

Common Reference

Architecture, programming models, and technical collateral that span every deployment.

Instinct Micro-architecture Hardware details for MI350, MI300, MI200, and MI100 accelerators. AMD SMI API Reference Full AMD SMI documentation covering all use cases. HIP C++ Learn the HIP programming model. OpenMP Explore the OpenMP programming model. Technical Information Portal NDA technical documentation and design collateral. Login required.

On this page