SPIRAL: Code Generation for DSP Transforms


Fast changing, increasingly complex, and diverse computing platforms pose central problems in scientific computing: How to achieve, with reasonable effort, portable optimal performance? We present SPIRAL that considers this problem for the performance-critical domain of linear digital signal processing (DSP) transforms. For a specified transform, SPIRAL automatically generates high performance code that is tuned to the given platform. SPIRAL formulates the tuning as an optimization problem, and exploits the domain-specific mathematical structure of transform algorithms to implement a feedback-driven optimizer. Similar to a human expert, for a specified transform, SPIRAL “intelligently” generates and explores algorithmic and implementation choices to find the best match to the computer’s microarchitecture. The “intelligence” is provided by search and learning techniques that exploit the structure of the algorithm and implementation space to guide the exploration and optimization. SPIRAL generates high performance code for a broad set of DSP transforms including the discrete Fourier transform, other trigonometric transforms, filter transforms, and discrete wavelet transforms. Experimental results show that the code generated by SPIRAL competes with, and sometimes outperforms, the best available human tuned transform library code.

DOI: 10.1109/JPROC.2004.840306

Extracted Key Phrases

30 Figures and Tables

Citations per Year

776 Citations

Semantic Scholar estimates that this publication has 776 citations based on the available data.

See our FAQ for additional information.

Cite this paper

@article{Pschel2005SPIRALCG, title={SPIRAL: Code Generation for DSP Transforms}, author={Markus P{\"{u}schel and Jos{\'e} M. F. Moura and Jeremy R. Johnson and David A. Padua and Manuela M. Veloso and Bryan Singer and Jianxin Xiong and Franz Franchetti and Aca Gacic and Yevgen Voronenko and Kang Chen and Robert W. Johnson and Nicholas Rizzolo}, journal={Proceedings of the IEEE}, year={2005}, volume={93}, pages={232-275} }