Title: Algorithms and analysis of scheduling for low-power high-performance DSP on VLIW processors

Authors: Zili Shao, Qingfeng Zhuge, Youtao Zhang, Edwin H-M. Sha

Addresses: Department of Computer Science, University of Texas at Dallas, Richardson, TX 75083, USA. ' Department of Computer Science, University of Texas at Dallas, Richardson, TX 75083, USA. ' Department of Computer Science, University of Texas at Dallas, Richardson, TX 75083, USA. ' Department of Computer Science, University of Texas at Dallas, Richardson, TX 75083, USA

Abstract: Switching activity and schedule length are the two most important factors that influence the energy consumption of an application executed on a VLIW (very long instruction word) processor. Considering these two factors together, we propose an instruction-level energy-minimisation scheduling technique to reduce the energy consumption of applications on VLIW processors. We first formally prove that this problem is NP-complete. Then three heuristic algorithms, MSAS, MLMSA, and EMSA, are proposed. While switching activity and schedule length are given higher priority in MSAS and MLMSA respectively, EMSA gives the best result considering both of them. The experimental results show that EMSA gives a 31.7% reduction in energy compared with the traditional list scheduling approach on average.

Keywords: DSP; VLIW processors; instruction level scheduling; low power consumption; very long instruction word architectures; energy minimisation; energy consumption; energy reduction; embedded systems; high performance computing; digital signal processing.

DOI: 10.1504/IJHPCN.2004.007561

International Journal of High Performance Computing and Networking, 2004 Vol.1 No.1/2/3, pp.4 - 16

Published online: 05 Aug 2005 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article