It is no secret that for some time, video traffic has risen almost off-the-scale. Back in 2016, Cisco’s Virtual Networking Index predicted that by 2020, 82% of all online traffic may well be video.
It is clear, as we all know that video is a big drain on bandwidth – and bandwidth is expensive, that to keep up with the increasing popularity, and of course, the evolution of the smartphone, tablet and other technologies, that greater investment is needed to ensure networks and operators can cope with increasing demand.
Another solution to the issue of diminishing bandwidth is advanced video compression, which can typically save up to 50% bandwidth demand. An example of historical success is the migration to high-definition from standard-definition video. However, many experts believe that due to the high number of mobile devices being utilised for video streaming, and at ever-increasing speeds, compression will not be enough.
With predictions at 82% video for all online traffic in fewer than three years away, experts believe that hardware acceleration will need to be implemented specifically for the video processing and encoding PRIOR to streaming.
Field-Programmable Gate Arrays (FPGAs) provide flexibility for video acceleration, and are predicted to accelerate AV1 coding by 10 times, when a comparison is made to software-based encoding running on a CPU. Multiple optimisation is allowed by programmable and reconfigurable capabilities of FPGAs across a broad spectrum of encoding profiles, as well as optimisation enablement for non-video processes.
AV1 codec, the first to be released last year by the Alliance for Open Media, replaced Google’s VP9 and directly competes with HEVC. This goes hand-in-hand with membership to the AOM, ensuring swift deployment of playback in OTT, smart TV, mobile devices and browsers.