Intel Corporation release of the library (Scalable Video Technology AV1), providing an alternative encoder and decoder for the AV1 video coding format, utilizing the hardware parallel computing capabilities present in modern Intel CPUs. The main goal of SVT-AV1 is to achieve a performance level suitable for real-time video transcoding and use in Video-On-Demand (VOD) services. The code is developed within the OpenVisualCloud project, which also develops encoders. and , and under the BSD license.
To use SVT-AV1, at least an Intel Core fifth generation processor (Intel Xeon E5-v4 and newer CPUs) is required. For encoding 10-bit AV1 streams at 4K quality, 48 GB of RAM is needed, for 1080p — 16 GB, for 720p — 8 GB, and for 480p — 4 GB. Due to the complexities of the algorithms used in AV1, encoding in this format requires significantly more resources than other formats, which prevents the standard AV1 encoder from being used for real-time transcoding. For instance, the standard encoder from the AV1 project
5721, 5869, and 658 times more computations compared to the x264 encoders (main profile), x264 (high profile), and libvpx-vp9.
Notable changes in the new release of SVT-AV1 include the addition of an initial decoder implementation (previously only encoding was supported) and support for static library linking. New SIMD optimizations have also been implemented, including support for using 2×2 filters, support for segmentation, and the ability for adaptive and adaptive INTRA block transformation.
Source: opennet.ru
