Based on Phoronix, (opens in new tab) Google has launched a brand new AOM-AV1 replace – model 3.5, that drastically improves encode occasions when streaming, rendering, or recording from the CPU. At its finest, the replace can enhance encoding occasions by as much as 34%.
It’s a unbelievable addition to AV1’s capabilities, with the encoder turning into extremely popular amongst highly effective video platforms akin to YouTube. As well as, we’re additionally seeing vital help for AV1 {hardware} acceleration on trendy discrete GPUs now, akin to Intel’s Arc Alchemist GPUs and, most significantly – Nvidia’s RTX 40-series GPUs.
Relying on the decision, encoding occasions with the brand new replace have improved by 20% to 30%. For instance, at 1080P, encode occasions that includes 16 threads of processing are decreased by 18% to 34%. At 4K, render occasions improved by 18% to twenty% with 32 threads. Google might do that by including Body Parallel Encoding to closely multi-threaded configurations. Google has additionally added a number of different enhancements contributing to AV1’s efficiency uplifts in different areas – particularly in real-time encoding.
In different phrases, CPU utilization in packages akin to OBS has been decreased, primarily for methods packing 16 CPU threads. In consequence, they’re permitting customers to make use of these CPU assets for different duties or improve video high quality even larger with none extra efficiency value. If you’re video modifying and are rendering out a video in AV1, processing occasions will probably be vastly decreased if in case you have a CPU with 16 threads or extra.
AV1’s Efficiency Enhancements Come At The Excellent Time
AV1’s pace enhancements could not have come at a greater time. The video business is aggressively transferring to AV1 greater than ever this 12 months, due to the introduction of AV1 {hardware} acceleration engines and permitting content material creators to make use of discrete GPUs to report and stream content material with the AV1 codec.
AV1 got here on the scene in 2018 as a more recent potential different to the H.264 video codec. It has develop into a mainstream merchandise resulting from its very enticing function set; for one, it’s totally open-source, permitting anybody to make use of it without spending a dime. It offers as much as 50% larger compression ratios than H.264. As well as, it’s drastically decreasing AV1 video file sizes.
This 12 months we noticed our first AV1 {hardware} acceleration engine within the Intel Arc A-series GPUs, offering drastic video high quality enhancements over its opponents, together with Nvidia’s legendary NVENC H.264 encoder. However now, Nvidia has lastly entered the sport with its RTX 40-series GPUs supporting AV1 encoding for the primary time – due to the brand-new eighth technology NVENC engine.
It units up the stage for AV1 to develop into a whole alternative for H.264, now that we have now AV1 encoding current in each discrete GPUs and by way of software program encoding on the CPU, which is able to make AV1 encoding very accessible to the general public. The brand new AV1 efficiency replace is simply the cherry on high and can present higher flexibility for customers working the AV1 codec on dwell streams and inside video editors.
