AMD ROCm™ documentation

Applies to Linux and Windows

2024-01-16

3 min read time

Welcome to the ROCm docs home page! If you’re new to ROCm, you can review the following resources to learn more about our products and what we support:

• What is ROCm?

• Release notes

Our documentation is organized into the following categories:

Installation

Installation guides

• Linux

  • Quick-start (Linux)

  • Linux install guide

  • Package manager integration

• Windows

  • Windows install guide

  • Application deployment guidelines

• Install Docker containers

• PyTorch for ROCm

• TensorFlow for ROCm

• MAGMA for ROCm

• ROCm & Spack

Compatibility & support

ROCm compatibility information

• System requirements (Linux)

• System requirements (Windows)

• Third-party

• User/kernel space

• Docker

• OpenMP

How-to

Task-oriented walkthroughs

• System tuning for various architectures

  • MI100

  • MI200

  • RDNA2

• Setting up for deep learning with ROCm

• GPU-enabled MPI

• System level debugging

• GitHub examples

Reference

Collated information

• API Libraries

Conceptual

Topic overviews & background information

• GPU architecture

  • MI100

  • MI200

  • MI250

• GPU memory

• Compiler disambiguation

• File structure (Linux FHS)

• GPU isolation techniques

• LLVM ASan

• Using CMake

• ROCm & PCIe atomics

• Inception v3 with PyTorch

• Inference optimization with MIGraphX

• OpenMP support in ROCm

We welcome collaboration! If you’d like to contribute to our documentation, you can find instructions on our Contribute to ROCm docs page. Known issues are listed on GitHub.

Licensing information for all ROCm components is listed on our Licensing page.