Home Content News Intel Arc GPUs Get AV1 Encoding Through Open Source Mesa

Intel Arc GPUs Get AV1 Encoding Through Open Source Mesa

0
1

Mesa’s open-source Intel ANV Vulkan driver now exposes Arc Alchemist’s dedicated AV1 hardware through a standard Vulkan API, expanding hardware video acceleration options for Linux applications.

Mesa’s open-source Intel ANV Vulkan driver has gained hardware-accelerated AV1 encoding through the VK_KHR_video_encode_av1 extension, exposing the dedicated AV1 encoding hardware in Intel’s DG2/Alchemist GPUs, including Arc A-Series graphics, to Linux applications through a standardised API.

The update does not add AV1 encoding to the GPUs; instead, it provides Vulkan applications with an open interface to hardware already present in the chips. This strengthens Mesa and Vulkan Video as an alternative to vendor-specific video acceleration APIs.

Finalised by the Khronos Group in 2024 and included in Vulkan 1.3.302, the AV1 encoding extension complements existing AV1 decoding and enables applications to query capabilities, configure encoding sessions, control rate settings, manage reference frames, and provide AV1-specific parameters.

With AV1 joining H.264 and H.265, Vulkan Video can now offer standardised interfaces for encoding and decoding all three codecs. The development follows July 2026 work that re-enabled H.264 and H.265 encoding for Intel Gen12.5/Alchemist hardware in ANV after insufficient testing had led to their earlier disablement.

The capability could benefit open-source multimedia projects such as FFmpeg and GStreamer, which are working on Vulkan Video support. However, applications must implement the functionality before users can access it.

AV1’s low-bitrate efficiency and royalty-free licensing make it relevant to streaming, recording, conferencing, broadcasting and transcoding. Contributions from developers associated with Red Hat and Igalia also underline the open-source development behind Mesa’s Vulkan Video stack.

The feature currently targets Alchemist hardware and requires supporting Mesa and applications, with wider GPU support dependent on further development and testing.

LEAVE A REPLY

Please enter your comment!
Please enter your name here