Intel oneVPL GPU Runtime 22.6.5 Brings AV1 & VP9 Encode Improvements
Following the recent release of the Intel Media Driver 2022Q4, Intel's oneVPL GPU runtime has been updated for its quarterly feature release that builds atop the Media Driver / VA-API stack and is about oneAPI integration for the video processing layer.
The oneVPL GPU Runtime 22.6.5 release, which is also labeled as part of the "2022Q4" release, has encode improvements on the AV1 and VP9 fronts. When it comes to AV1 encode there is reset functionality introduced, TCBRC support for Transport Controlled BRC Specific bitrate control for real time streaming, and a memory leak fix. On the oneVPL VP9 video encode front is a fix for an issue around the segmentation map alignment.
The Intel oenVPL GPU Runtime update also fixes H.264 Decoder Picture Buffer (DPB) management, optimizes the memory footprint around AV1 film grain handling, and improves the JPEG height alignment.
Downloads and more details on the Intel oneVPL GPU Runtime update via GitHub.
The oneVPL GPU Runtime implements oneAPI's oneVPL API for Intel Gen GPUs for Tigerlake and newer, plus the latest discrete graphics support. This runtime supports dealing with HEVC / AVC / MPEG-2 / JPEG / VP9 encoding and the same on the decode front plus AV1, VC1, and VP8. Various video pre-processing filters are also supported. This ultimately is the successor to Intel's Media SDK.
The oneVPL GPU Runtime 22.6.5 release, which is also labeled as part of the "2022Q4" release, has encode improvements on the AV1 and VP9 fronts. When it comes to AV1 encode there is reset functionality introduced, TCBRC support for Transport Controlled BRC Specific bitrate control for real time streaming, and a memory leak fix. On the oneVPL VP9 video encode front is a fix for an issue around the segmentation map alignment.
The Intel oenVPL GPU Runtime update also fixes H.264 Decoder Picture Buffer (DPB) management, optimizes the memory footprint around AV1 film grain handling, and improves the JPEG height alignment.
Downloads and more details on the Intel oneVPL GPU Runtime update via GitHub.
oneVPL-intel-gpu is just one of several oneVPL back-ends available.
The oneVPL GPU Runtime implements oneAPI's oneVPL API for Intel Gen GPUs for Tigerlake and newer, plus the latest discrete graphics support. This runtime supports dealing with HEVC / AVC / MPEG-2 / JPEG / VP9 encoding and the same on the decode front plus AV1, VC1, and VP8. Various video pre-processing filters are also supported. This ultimately is the successor to Intel's Media SDK.
1 Comment