Mechanical derivation of fused multiply-add algorithms for linear transforms

IEEE Transactions on Signal Processing

Volumn 55, Issue 9, 2007, Pages 4458-4473

  Voronenko, Yevgen ^a   Püschel, Markus ^a  

^a CARNEGIE MELLON UNIVERSITY   (United States)

Author keywords

Automatic program generation; Discrete cosine transform (DCT); Discrete Fourier transform (DFT); Fast algorithm; Implementation; Multiply and accumulate (MAC); Multiply and accumulate (MAC) instruction

Indexed keywords

AUTOMATIC PROGRAMMING; DISCRETE COSINE TRANSFORMS; DISCRETE FOURIER TRANSFORMS; MATRIX ALGEBRA; OPTIMIZATION;

DIRECTED ACYCLIC GRAPHS; FUSED MULTIPLY-ADD; LINEAR TRANSFORM; MULTIPLY-AND-ACCUMULATE INSTRUCTION;

COMPUTER ARCHITECTURE;

EID: 34548317511     PISSN: 1053587X     EISSN: None     Source Type: Journal    
DOI: 10.1109/TSP.2007.896116     Document Type: Article

References (26)

1. Y. Nievergelt, "Scalar fused multiply-add instructions produce floating-point matrix arithmetic provably accurate to the penultimate digit," ACM Trans. Math. Softw. (TOMS), vol. 29, no. 1, pp. 27-48, 2003.
2. E. Linzer and E. Feig, "Implementation of efficient FFT algorithms on fused multiply-add architectures," IEEE Trans. Signal Process., vol. 41, no. 1, p. 93, Jan. 1993.
3. C. Lu, "Implementation of multiply-add FFT algorithms for complex and real data sequences," in Proc. Int. Symp. Circuits Systems (ISCAS) 1991, vol. 1, pp. 480-483.
4. S. Goedecker, "Fast radix 2, 3, 4, and 5 kernels for fast Fourier transformations on computers with overlapping multiply-add instructions," SIAM J. Scientif. Comput., vol. 18, no. 6, pp. 1605-1611, 1997.
5. C. Lu, J. W. Cooley, and R. Tolimieri, "FFT algorithms for prime transform sizes and their implementations on VAX, IBM3090VF, and IBM RS/ 6000," IEEE Trans. Signal Process., vol. 41, no. 2, pp. 638-648, Feb. 1993.
6. D. Takahashi, "A new radix-6 FFT algorithm suitable for multiply-add instruction," in Proc. IEEE Int. Conf. Acoustics, Speech, Signal Processing (ICASSP), Jun. 2000, vol. 6, pp. 3343-3346.
7. D. Takahashi, "A radix-16 FFT algorithm suitable for multiply-add instruction based on Goedecker method," in Proc. IEEE Int. Conf. Acoustics, Speech, Signal Processing (ICASSP), Apr. 2003, vol. 2, pp. 665-668.
8. E. Linzer and E. Feig, "New scaled DCT algorithms for fused multiply/ add architectures," in Proc. IEEE Int. Conf. Acoustics, Speech, Signal Processing (ICASSP), 1991, vol. 3, pp. 2201-2204.
9. E. F. D. Coppersmith and E. Linzer, "Hadamard transforms on multiply/ add architectures," IEEE Trans. Signal Process., vol. 42, no. 4, pp. 969-970, Apr. 1994.
10. Y. Voronenko and M. Püschel, "Automatic generation of implementations for DSP transforms on fused multiply-add architectures," in Proc. IEEE Int. Conf. Acoustics, Speech, Signal Processing (ICASSP), 2004, vol. 5, pp. 101-104.
11. M. Püschel, J. M. F. Moura, J. Johnson, D. Padua, M. Veloso, B. W. Singer, J. Xiong, F. Franchetti, A. Gačić, Y. Voronenko, K. Chen, R. W. Johnson, and N. Rizzolo, "SPIRAL: Code generation for DSP transforms," Proc. IEEE (Special Issue on Program Generation, Optimization, and Adaptation), vol. 93, no. 2, pp. 232-275, 2005.
12. Spiral website [Online]. Available: Www.spiral.net , "," vol. , pp. -,
13. P. Bürgisser,M. Clausen, and M. A. Shokrollahi, Algebraic Complexity Theory. New York: Springer, 1997.
14. C. Van Loan, Computational Framework of the Fast Fourier Transform Philadelphia, PA: SIAM, 1992.
15. N. Dershowitz and D. A. Plaisted, "Rewriting," in Handbook of Automated Reasoning, A. Robinson and A. Voronkov, Eds. New York: Elsevier, 2001, vol. 1, ch. 9, pp. 535-610.
16. F. Franchetti, Y. Voronenko, and M. Püschel, "Loop merging for signal transforms," in Proc. Programming Language Design Implementation (PLDI), 2005, pp. 315-326.
17. F. Franchetti and M. Püschel, "A SIMD vectorizing compiler for digital signal processing algorithms," in Proc. IEEE Int. Parallel Distributed Processing Symposium (IPDPS), 2002, pp. 20-26.
18. F. Franchetti, Y.Voronenko, and M. Püschel, "FFT program generation for shared memory: SMP and multicore," in Proc. Supercomputing, 2006.
19. FFTW 3.1.2 2006 [Online]. Available: Www.fftw.org
20. M. Frigo and S. G. Johnson, "FFTW: An adaptive software architecture for the FFT," in Proc. IEEE Int. Conf. Acoustics, Speech, Signal Processing (ICASSP), 1998, vol. 3, pp. 1381-1384 [Online]. Available: www.fftw.org
21. M. Frigo, "A fast Fourier transform compiler," in Proc. ACM SIGPLAN Conf. Programming Language Design Implementation (PLDI), 1999, pp. 169-180.
22. H. V. Sorensen, H. M. T. , C. S. Burrus, and M. T. Heideman, "On computing the split-radix FFT," IEEE Trans. Acoust., Speech, Signal Process., vol. ASSP-34, no. 1, pp. 152-156, 1986.
23. K. R. Rao and P. Yip, Discrete Cosine Transform: Algorithms, Advantages, Applications. New York: Academic, 1990.
24. F. Bruekers and A. Enden, "New networks for perfect inversion and perfect reconstruction," IEEE J. Sel. Areas Commun., vol. 10, pp. 130-137, Jan. 1992.
25. W. Sweldens, "The lifting scheme: A custom-design construction of biorthogonal wavelets," Appl. Comput. Harmon. Anal., vol. 3, no. 2, pp. 186-200, 1996.
26. K. Serebryany, "Optimal placement of fused multiply-add (FMA) instructions," presented at the 6th Workshop on Explicitly Parallel Instructions, Computing Architectures and Compiler Technology (EPIC-6), San Jose, CA, Mar. 2007.