Strategy for power ecient combined task and data parallelism exploration illustrated on a QSDPCM video codec Koen Danckaert a, * ,1 , Kostas Masselos b , Francky Catthoor a,2 , Hugo De Man a,2 a IMEC, Kapeldreef 75, B-3001 Leuven, Belgium b Department of Electrical and Computer Engineering, University of Patras, Rio 26500, Greece Abstract Application studies indicate that between 50% and 80% of the power cost in image and video processing systems is due to data storage and transfers. This is especially true for multi-processor realizations, because conventional par- allelization strategies ignore this cost and focus only on the performance, whereas the power consumption also depends heavily on the way a system is parallelized. We will demonstrate the impact of processor partitioning on the memory requirements by exploring a QSDPCM video codec realization. Furthermore, we show that a strategy for combined task and data parallelism exploration leads to a signi®cant power reduction. Ó 1999 Elsevier Science B.V. All rights reserved. Keywords: Memory optimization; Low-power design; Multi-media; Processor partitioning; Hybrid task-data parallelism 1. Introduction and related work In multi-media applications and others that make use of large multi-dimensional array-type data structures, a very large amount of memory is required. In embedded applications, which we target in this paper, memory forms the main part of the total area and power cost of the system [13,23]. Power consumption is of course especially a problem for mobile applications, where battery life is one of the main design issues; but also packaging cost reduction and especially reliability improvements require a reduction of the current levels of chip and board power consumption. For Journal of Systems Architecture 45 (1999) 791±808 * Corresponding author. E-mail: danckaer@imec.be 1 Research Assistant of the Fund for Scienti®c Research ± Flanders. 2 Also professor at the Katholieke Universiteit Leuven, Belgium. 1383-7621/99/$ ± see front matter Ó 1999 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 7 6 2 1 ( 9 8 ) 0 0 0 3 9 - 3