Merged last month to the widely-used FFmpeg open-source multimedia library was an initial Vulkan-based decoder for FFV1 for the FF Video 1 lossless video coding format. Should you be interested in using this FFV1 decoder on AMD GPUs, there’s a reported 3x performance improvement with the newest code.
Prominent FFmpeg developer Lynne who has been working on many of the project’s Vulkan Video initiatives merged a lot of “vulkan_ffv1” improvements this week. Among the work is now enabling FFV1 acceleration on Intel hardware. There is also various optimizations.
The work most interestingly of this recent batch culminates with vulkan_ffv1: add cached symbol reader for AMD. That Git commit message speaks for itself:
“vulkan_ffv1: add cached symbol reader for AMD
Speeds up everything on AMD by 3x.
This uses 32 local invocations to load state into cache, as well as to do the RCT faster.”
It’s quite an improvement for those wanting to Vulkan acceleration for FFV1 videos on AMD Radeon GPUs.
