Welcome to Intel® Gaudi® v1.16 Documentation

Find detailed documentation to learn how to use the Intel Gaudi solutions - first-gen Intel® Gaudi® AI accelerator and Intel® Gaudi® 2 AI accelerator. This will cover the details on how to migrate models to Gaudi, code samples, diagrams, best practices for debug and optimization, API references, and more.

Please navigate the Table of Contents on the left or use the options below for the most popular activities:

Get Access to Gaudi Instances

Connect to the Intel Developer Cloud for Gaudi 2 or a DL1 instance for first-gen Gaudi

Use the Installation Guide to set up Software and System

Get Started with PyTorch

Run Generative AI or Large Language examples with DeepSpeed

Run PyTorch simple models

Run Hugging Face examples

Get Functional

Migrate a model to Gaudi using the GPU Migration Tool

Use the Intel Gaudi Docker Image in your instance

See the PyTorch page for additional features

Get Optimized

Follow the Optimization Guide to get your model optimized for performance

IT Admin

Set up the platform for a full platform system installation, including FW management, SW installation, platform setup and verification

Model Catalog

Use the Model Catalog to start with PyTorch models already running on Gaudi