AMD Xilinx Video SDK 3.0 brings additional enhancements to the already feature-rich Video SDK 2.0. It is now available to Alveo™ U30 media accelerator card users through on-premises setups (support by AWS’s Amazon EC2 VT1 instances coming soon). With Video SDK 3.0 and a multimedia framework/toolkit (such as GStreamer, FFmpeg, and XMA), users can continue to cost-effectively scale their live video transcoding with even greater ease.
In case you haven’t heard, the AMD Xilinx Video SDK 3.0 has been released and will be supported by AWS’s Amazon EC2 VT1 Instances shortly. For Alveo U30 media accelerator card users who need to handle growing volume of video transcoding workloads (such as live streaming, video conferencing, video library optimization, and just-in-time asset transcoding), Video SDK 3.0 adds a number of enhancements (see below) to Video SDK 2.0 (which already supports 10-bit color, HDR10/HDR 10+, and lookahead for all resolutions amongst other things).
- Ultra low latency (ULL) encoding
- Dynamic GOP
- Min/max frame quantization parameter (QP) bounding
These enhancements are applicable to GStreamer, FFmpeg, and XMA.
What this means:
For live streaming applications such as video collaboration and cloud gaming where every millisecond counts, subframe encode latency is now available as an option in the ULL mode.
Additionally, with the new dynamic GOP support and on-the-fly min/max QP adjustment, users now have the ability to optimize the bitrate to the available bandwidth.
Finally, this release brings with it enhanced OS/kernel support (such as Ubuntu kernel 5.15 and Amazon Linux 2 kernel 5.10) and bug fixes.
Visit What's New (Xilinx Video SDK Release Notes — Xilinx Video SDK 3.0 (Production) documentation) for the full list of enhancements of Video SDK 3.0. (For that of Video SDK 2.0, please visit here (Xilinx Video SDK Release Notes — Xilinx Video SDK 2.0.1 (Production) documentation).)