.. meta:: :description: Central hub for AMD Instinct™ cluster design and deployment documentation. :keywords: network validation, DCGPU, PCIe, Infiniband, RoCE, ROCm, RCCL, machine learning, LLM, usage, tutorial :html_theme.sidebar_secondary.remove: true .. raw:: html

AMD Instinct™ — Cluster Design and Deployment

************************************************************************************************************************ Cluster Documentation Hub ************************************************************************************************************************ .. raw:: html .. raw:: html

Start here

Design and Guides

Configure, validate, and optimize AMD Instinct™-based clusters.

Instinct Customer Acceptance Guide Comprehensive guide for configuring, validating, benchmarking, and baselining AMD Instinct™-based systems. Cluster Networking Guide Optimize the network for AMD Instinct™ GPU applications. MI3XX Reference Design Reference design materials and topology examples for the AMD Instinct™ MI3XX platform. Data Center Design Guide How to plan and design the modern AI data center. AMD-DriveNets System Reference Architecture Validated, end-to-end reference architecture for building and operating large-scale AI GPU clusters using AMD Instinct MI350-series (MI355X) compute paired with the DriveNets AI Fabric networking solution.

.. raw:: html

Reference

Articles and Overviews

Technical analyses and background reading that span AMD Instinct™ products.

Comparative Analysis of Scale-Out RoCE Network Traffic Patterns and Loads in Training Large Language Models Compares RoCE network traffic patterns and loads across GPT-4, Llama 3, DeepSeek-V2, and Grok 4.0 LLM training. AMD Instinct™ Network Traffic, Congestion Trends, and Harmonics in Scale-Out Networks for AI Training Clusters Explore how synchronized GPU collectives create harmonic congestion in AI clusters and the strategies to diagnose and mitigate it.