Video compression plays a central role in a vast number of multimedia appli
cations but its computational requirements overwhelm the capabilities of an
y present single processor system. In this paper, we explore the use of par
allel machines like the Intel Paragon to compress MPEG-2 video sequences. T
he motivation is to build a production-based compression facility by exploi
ting the potential power of the available machine. Given a video sequence o
r a set of sequences, the aim of the parallel encoder is to achieve the max
imum possible encoding rate. A collective scheduling scheme for the process
ors, I/O nodes, and disks is proposed that provides fast I/O, minimizes the
idle times of processors, and enables the system to work in a highly balan
ced fashion. An efficient data layout scheme for storing video frames is al
so proposed in order for the I/O to sustain the desired data transfer rates
. Using a small percentage of processors as the I/O nodes results in an eff
icient utilization of the system resources. As shown by experimental and an
alytical results, the encoding scheme is scalable and higher performance ca
n be achieved with larger machines. The performance of the proposed scheme
can be many times the real-time encoding rates with Standard Interface Form
at (SIF) and CCIR-601 video sequences. The experimental results indicate ab
out two-fold gain in performance compared to the previous studies. Such a s
ystem is useful for the conversion of analog videos to compressed digital f
orm in large studios, digital libraries, and other multimedia database envi
ronments. The proposed scheme partitions the system into groups of compute
nodes, and I/O nodes, and can be easily extended to other MIMD machines or
a set of networked workstations. (C) 2001 Elsevier Science B.V. All rights
reserved.