GCC 14 Adds "GFX90C" For OpenMP Offloading To APUs With GFX9/Vega Graphics

As the last feature patch prior to the GCC 14 compiler code being branched today and GCC 15 opening up on the mainline codebase, AMD GFX90C support was merged for enabling GPU OpenMP device offloading to the numerous AMD SoCs/APUs with the GFX9/Vega graphics.

Frederik Harwath contributed the patch for enabling the AMD GFX90C graphics to be supported alongside the other GFX9 / GFX10 / GFX11 graphics supported by the GNU Compiler Collection. The AMD GCN target within the GCC compiler is geared just for OpenMP device offloading support with the AMDGPU LLVM back-end otherwise what's being commonly used for numerous uses inside and outside of AMD. The AMD GCN back-end for GCC doesn't receive too much attention outside of various external contributors working on OpenMP and OpenACC offloading.

Harwath explained in the newly-merged patch:

"Add support for gfx90c GCN5 APU integrated graphics devices. The LLVM AMDGPU documentation does not list those devices as supported by rocm-amdhsa, but it passes most libgomp offloading tests. Although they are constrainted compared to dGPUs, they might be interesting for learning, experimentation, and testing."

With the GFX90C APU integrated graphics being very common on prior generation Ryzen systems, this could lead to some interesting uses and at least opening the door for more developers to experiment with GCC device offloading.

This GFX90C support was merged just in time for the GCC 14 branching and thus will be part of the GCC 14.1 stable release coming up in a few weeks. GCC 14 on the AMD side also adds the new Zen 5 "znver5" CPU target and the RDNA2 and RDNA3 GPU offloading support.