Performance of software-based MPEG-2 video encoder on parallel and distributed systems

Abstract

Video encoding due to its high processing requirements has been traditionally done using special-purpose hardware. Software solutions have been explored but are considered to be feasible only for nonreal-time applications requiring low encoding rates. However, a software solution using a general-purpose computing system has numerous advantages: it is more available and flexible and allows experimenting with and hence improving various components of the encoder. In this paper, we present the performance of a software video encoder with MPEG-2 quality on various parallel and distributed platforms. The platforms include an Intel Paragon XP/S and an Intel iPSC/860 hypercube parallel computer as well as various networked clusters of workstations. Our encoder is portable across these platforms and uses a data-parallel approach in which parallelism is achieved by distributing each frame across the processors. The encoder is useful for both real-time and nonreal-time applications, and its performance scales according to the available number of processors. In addition, the encoder provides control over various parameters such as the size of the motion search window, buffer management, and bit rate. The performance results include comparisons of execution times, speedups, and frame encoding rates on various systems.