OpenACC


This page contains information on GCC's implementation of the OpenACC specification and related functionality. OpenACC is intended for programming accelerator devices such as GPUs, including code offloading to these devices. Currently devices using Nvidia PTX (nvptx) and AMD GCN devices (Fiji and Vega GPUs).

Using OpenACC

Use the -fopenacc flag to enable OpenACC compilation; see compilation options for additional flags.

Active GCC Releases

Development Branch

The og10 development branch is known as devel/omp/gcc-10 in the GCC git repository.

Resources

This branch implements most of the OpenACC 2.6 specification and contains further improvements. Some features available on the development branch are not present on the GCC release branch. (More details on og10, and GCC 10.)

For discussing this project, please use the standard GCC resources (mailing lists, Bugzilla, and so on). It's helpful to put a [OpenACC] tag into your emails' subject line, and set the openacc keyword in any Bugzilla issues filed.

For instructions how to build offloading-enabled GCC from source, see Offloading: How to try offloading enabled GCC.

Technical Details

Issue Tracking

Open OpenACC bugs

None: OpenACC (last edited 2020-09-24 06:46:02 by TobiasBurnus)