Methodology For Behavioral Synthesis-based Algorithm-level Design Space Exploration: DCT Case Study

Abstract

Numerous fast algorithms for the Discrete Cosine Transform (DCT) have been proposed. Until recently, it has been difficult to compare different DCT algorithms and select one which is best suited for implementation under a given set of design goals. We propose an approach for design space exploration at the algorithm and behavioral levels using behavioral synthesis tools and demonstrate its effectiveness for designing DCT ASIC. In particular, we study and compare the following nine DCT algorithms: Lee's, Wang's, DIT, DFT, Arai's, DIF, Vetterli's, planar rotation, and direct algorithm. The main conclusions of this study are (i) the best choice among fast DCT algorithms depends on a particular set of design goals and constraints and (ii) almost always more than an order of magnitude improvement can be achieved using algorithm and behavioral design space exploration.

Advanced Techniques for Embedded Systems Design and Test

“…Similarly, at the algorithmic level, one may attempt to reduce power consumption using, for example, operation strength reduction on power hungry operations, such as multiplication. A recent case study illustrates the results of applying one of such transformations --specifically, "multiplications" substituted by "adds" and "shifts" --to fourteen DCT algorithms, considering standard-cell implementations [PoKi97]. Interestingly, in all cases the overall power consumption has actually increased as a result of the transformation --in half of the cases it actually increased by more than 40%.…”