Benchmarking Performances of Raspberry pi Microcomputer as Video Wall Devices

Video wall development is affected by cost, power consumption, processing capabilities, algorithm and video used. Literature has shown that using microcomputers reduces power consumption and cost, but performance remains a bottleneck. Benchmarking the performances of Raspberry pi (R-pi) devices with real-world loads will help understand R-pi video wall development, suitability and utilization. The approach used in this work is based on parallel video streaming using user datagram protocols (UDP) and broadcast addressing, while image splitting is done on clients. Nigel's performance monitoring (NMON) tool was used with videos of varying frames 15fps, 20fps, 24fps, 25fps, 30fps, 50fps and 60fps and resolutions of 144p, 240p, 360p, 480p, 720p and 1080p to benchmark performances. Results revealed a maximum of 9.78%, 17.16%, 58.45 kB/s, and 1.13 kB/s for central processing unit (CPU), memory, network, and disk usage, respectively. Results also reveal that R-pi as a video wall device with the proposed approach has the processing capability towards enhancing video wall development. These results reveal for best performances, R-pi video walls are more suitable with videos of higher resolutions such as 480p, 720p and 1080p and at lower frame rates such as 24fps, 25fps and 30fps.