When you purchase through links on our site, we may earn an affiliate commission. Here’s how it works.

Intel and Netflix's SVT-AV1 gains up to 40% performance boost, AMD Ryzen will benefit too

Intel Arc AV1
Intel Arc became the first GPU with full AV1 support

Back in 2019, Intel and Netflix collaborated on creating a software-based (CPU) encoding implementation of the AV1 codec. As such, SVT-AV1 was born where SVT stands for Scalable Video Technology. SVT was developed by the former, and as the name suggests, it is designed to be highly scalable in order to get the best out of multi-threaded processors and multi-socket CPU systems. And while SVT was developed by Intel, AMD's Ryzen and EPYC chips also do quite well in SVT-AV1 benchmarks.

Following that, SVT-AV1 was adopted by the Alliance for Open Media (AOMedia) as well and the latest version of the codec brings a massive performance boost of up to 40% under certain situations. The official release notes claim a 30-40% speed up when encoding in higher-quality presets. However, it is unknown if the testing was done on Intel or AMD or an average of both.

Meanwhile, the faster (lower fidelity) presets have a modest 1-4% improvement as well in their Bjontegaard delta rate (BD rate). For those unaware, the BD rate helps to measure and compare the compression efficiency of various codecs.

The full changelog of the SVT-AV1 version 1.6.0 is given below:

Encoder

  • Improve the tradeoffs for the random access mode across presets M1-M13:
  • Speeding up the higher quality presets by 30-40%
  • Improving the BD-rate by 1-4% for the faster presets
  • Improve the tradeoffs for the low delay mode for both screen content and non-screen content encoding modes
  • Add a toggle to remove the legacy one-frame buffer at the input of the pipeline allowing the low delay mode to operate at sub-frame processing latencies
  • Add a new API allowing the user to specify quantization offsets for a region of interest per frame

Build, cleanup and bug fixes

  • Various cleanups and functional bug fixes
  • Fix the startup minigop size BD-rate loss
  • Add ability to run the ci-testing offline

In related news, preliminary support for the next-gen AV2 codec is already beginning to surface.

Report a problem with article
snap generative ai art
Next Article

Making a generative AI image might take less than two seconds on a smartphone in the future

The Apple logo
Previous Article

Apple is forcing a 111-year old fruit company to change its logo in a weird trademark battle