HIP documentation

The Heterogeneous-computing Interface for Portability (HIP) is a C++ runtime API and kernel language that lets you create portable applications for AMD and NVIDIA GPUs from a single source code. For more information, see What is HIP?

Installation instructions are available from:

• Installing HIP

• Building HIP from source

HIP enabled GPUs:

• Supported AMD GPUs on Linux

• Supported AMD GPUs on Windows

• Supported NVIDIA GPUs

The HIP documentation is organized into the following categories:

Conceptual

• HIP programming model

• Hardware implementation

• AMD common language runtimes (CLR)

• Texture fetching

How to

• Programming manual

• HIP porting guide

• HIP porting: driver API guide

• Programming for HIP runtime compiler (RTC)

• Performance guidelines

• Debugging with HIP

• Logging HIP activity

• Unified memory

• Virtual memory

• Stream Ordered Memory Allocator

• Cooperative groups

• HIP graphs

Reference

• HIP runtime API

  • Modules

  • Global defines, enums, structs and files

• HSA runtime API for ROCm

• C++ language extensions

• C++ language support

• HIP math API

• HIP environment variables

• Comparing syntax for different APIs

• List of deprecated APIs

• FP8 numbers in HIP

Tutorial

• HIP basic examples

• HIP examples

• HIP test samples

• SAXPY tutorial

• Reduction tutorial

• Cooperative groups tutorial

Known issues are listed on the HIP GitHub repository.

To contribute features or functions to the HIP project, refer to Contributing to HIP. To contribute to the documentation, refer to Contributing to ROCm docs page.

You can find licensing information on the Licensing page.