School of Computer Science Intranet
Ahmed El-Mahdy and Ian Watson
Vector processing is gaining attention for supp orting multi- media workloads, particularly small subword vectors. In this pap er we prop ose a novel vector instruction set combining the b enefits of subword parallelism and traditional vector processing. We also develop a simple cache prefetching optimisation that exploits the two dimensional data ac- cess pattern of multimedia MPEG2 video applications. The architecture parameter space is explored by a simple analytical study. The analysis is complemented by detailed simulation of the actual system where it is shown that the optimised cache removes 75% of the misses and the instruction set p erformance is equivalent to a subword instruction with double the word size on the average.
(LNCS 2150). Springer, 2001. (© Springer-Verlag)