Coding Efficiency and Time Evaluation of Apple A15 Bionic Chipset HEVC Encoder
Vı́tor Santos Costa, Murilo Perleberg, Luciano Volcan Agostini, Marcelo Porto · 2024
Current mobile devices need to be equipped with dedicated hardware for video encoding/decoding to achieve high performance and to be capable of real-time processing of high-resolution videos. In dedicated hardware video coding, various constraints must be adopted in order to achieve efficient encoding with regard to time and energy consumption. Such constrains involves to remove (or to simplify) many complex tools to speed up the encoding process and/or to reduce the area/memory/power demand. This work evaluates the impact of such restrictions applied to the dedicated hardware of High Efficiency Video Coding (HEVC) video encoder of the chipset Apple A15 Bionic, embedded in the iPhone 13 and iPhone 14 smartphones. The coding efficiency and execution time of the HEVC encoder were evaluated in comparison with the reference software implementation of the HEVC standard for HD 1080p and UHD 4K videos. The results in this paper offer reference values for coding-efficiency losses and time reduction achieved by a state-of-the-art commercial dedicated HEVC encoder chip. These values can serve as benchmark parameters for evaluating the coding efficiency of other dedicated HEVC chips.