Is search really necessary to generate high-performance BLAS?

Proceedings of the IEEE

Volumn 93, Issue 2, 2005, Pages 358-385

  Yotov, Kamen ^a   Li, Xiaoming ^b   Ren, Gang ^b   Garzarán, María Jesús ^b   Padua, David ^b   Pingali, Keshav ^a   Stodghill, Paul ^a  

^a School of Operations Research and Information Engineering   (United States)

^b University of Illinois at Urbana Champaign   (United States)

Author keywords

Basic Linear Algebra Subprograms (BLAS); Compilers; Empirical optimization; High performance computing; Library generators; Model driven optimization; Program optimization

Indexed keywords

COMPUTER OPERATING SYSTEMS; COMPUTER SOFTWARE; MATHEMATICAL TRANSFORMATIONS; OPTIMIZATION; PARAMETER ESTIMATION; PROGRAM COMPILERS;

BASIC LINEAR ALGEBRA SUBPROGRAMS; EMPIRICAL OPTIMIZATION; HIGH PERFORMANCE COMPUTING; LIBRARY GENERATORS; MODEL-DRIVEN OPTIMIZATION; PROGRAM OPTIMIZATION;

SEARCH ENGINES;

EID: 20744459570     PISSN: 00189219     EISSN: None     Source Type: Journal    
DOI: 10.1109/JPROC.2004.840444     Document Type: Conference Paper

References (43)

1. ATLAS home page [Online]. Available: http.//math-atlas.sourceforge.net/
2. PHiPAC home page [Online]. Available: http://www.icsi.berkeley.edu/ ~bilmes/phipac
3. R. C. Agarwal, F. G. Gustavson, and M. Zubair, "Improving performance of linear algebra algorithms for dense matrices using algorithmic prefetch," IBM J. Res. Develop., vol. 38, no. 3, pp. 265-275, 1994.
4. R. Allan and K. Kennedy, Optimizing Compilers for Modern Architectures. San Francisco, CA: Morgan Kaufmann, 2002.
5. U. Banerjee, "Unimodular transformations of double loops," in Advances in Languages and Compilers for Parallel Processing. Cambridge, MA: MIT Press, 1991, ch. 10, pp. 192-219.
6. J. Bilmes, K. Asanović, C.-w. Chin, and J. Demmel, "Optimizing matrix multiply using PHiPAC: A portable, high-performance, ANSI C coding methodology," presented at the Int. Conf. Supercomputing, Vienna, Austria, 1997.
7. P. Boulet, A. Darte, T. Risset, and Y. Robert, "(Pen)-ultimate tiling?," Integration VLSI J., vol. 17, pp. 33-51, 1994.
8. D. Callahan, J. Cocke, and K. Kennedy, "Estimating interlock and improving balance for pipelined architectures," J. Parallel Distrib. Comput., vol. 5, no. 4, pp. 334-358, 1988.
9. D. Callahan, S. Carr, and K. Kennedy, "Improving register allocation for subscripted variables," in Proc. SIGPLAN Conf. Programming Language Design and Implementation, 1990, pp. 53-65.
10. S. Chatterjee, E. Parker, P. J. Hanlon, and A. R. Lebeck, "Exact analysis of the cache behavior of nested loops," in Proc. ACM SIGPLAN 2001 Conf. Programming Language Design and Implementation, 2001, pp. 286-297.
11. M. Cierniak and W. Li, "Unifying data and control transformations for distributed shared memory machines," in SIGPLAN 1995 Conf. Programming Languages Design and Implementation, pp. 205-217.
12. S. Coleman and K. S. McKinley, "Tile size selection using cache organization and data layout," in Proc. SIGPLAN Conf. Programming Language Design and Implementation, 1995, pp. 279-290.
13. P. Feautrier, "Some efficient solutions to the affine scheduling problem - Part 1: One dimensional time," Int. J. Parallel Program., vol. 1, no. 5, pp. 313-348, Oct. 1992.
14. M. Fowler, "Yet another optimization article," IEEE Softw., vol. 19, no. 3, pp. 20-21, May/Jun. 2002.
15. B. B. Fraguela, R. Doallo, and E. Zapata, "Automatic analytical modeling for the estimation of cache misses," in Proc. Int. Conf. Parallel Architectures and Compilation Techniques (PACT), 1999, pp. 221-231.
16. M. Frigo and S. G. Johnson, "FFTW: An adaptive software architecture for the FFT," in Proc. IEEE Intl. Conf. Acoustics, Speech, and Signal Processing, vol. 3, 1998, pp. 1381-1384.
17. _, "The design and implementation of FFTW3," Proc. IEEE, vol. 93, no. 2, pp. 216-231, Feb. 2005.
18. S. Goedecker and A. Hoisie, Performance Optimization of Numerically Intensive Codes. Philadelphia, PA: SIAM, 2001.
19. K. Goto and R. van de Geijn, "On reducing TLB misses in matrix multiplication," Dept. Comput. Sci., Univ. Texas, Austin, Tech. Rep. TR-2002-55, 2002.
20. J. A. Gunnels, G. M. Henry, and R. A. van de Geijn, "A family of high-performance matrix algorithms," in Proc. Int. Conf, Computational Science (ICCS 2001), pp. 51-60.
21. F. Gustavson, private communication, 2004.
22. J. L. Hennessy and D. A. Patterson, Computer Architecture: A Quantitative Approach. San Francisco, CA: Morgan Kaufmann, 1990.
23. "Flexible High-Performance Matrix Multiply via Self-Modifying Runtime Code," Dept. Comput. Sci., Univ. Texas, Austin, Tech. Rep. CS-TR-01-44, Dec. 2001.
24. K. Goto. High-performance BLAS. [Online]. Available: http://www.cs. utexas.edu/users/flame/goto/
25. J. Johnson, R. W. Johnson, D. A. Padua, and J. Xiong, "Searching for the best FFT formulas with the SPL compiler," in Proc. 13th Int. Workshop on Languages and Compilers for Parallel Computing, 2000, pp. 109-124.
26. I. Kodukula, N. Ahmed, and K. Pingali, "Data-centric multi-level blocking," in Proc. ACM SIGPLAN Conf. Programming Language Design and Implementation, 1997, pp. 346-357.
27. I. Kodukula and K. Pingali, "Imperfectly nested loop transformations for memory hierarchy management," presented at the Int. Conf. Supercomputing, Rhodes, Greece, June 1999.
28. W. Li and K. Pingali, "Access normalization: Loop restructuring for NUMA compilers," ACM Trans. Comput. Syst., 1993.
29. R. L. Mattson, J. Gecsei, D. R. Slutz, and I. L. Traiger, "Evaluation techniques for storage hierarchies," IBM Syst. J., vol. 9, no. 2, pp. 78-92, 1970.
30. A. C. McKellar and E. G. Coffman Jr., "Organizing matrices and matrix operations for paged memory systems," Commun. ACM, vol. 12, no. 3, pp. 153-165, 1969.
31. D. Padua and M. Wolfe, "Advanced compiler optimization for supercomputers," Commun. ACM, vol. 29, no. 12, pp. 1184-1201, Dec, 1986.
32. 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, vol. 93, no. 2, pp. 232-275, Feb. 2005.
33. J. McComb, R. C. Agarwal, F. G. Gustavson, and S. Schmidt, "Engineering and scientific subroutine library release 3 for IBM ES/3090 vector multiprocessors," IBM Syst. J., vol. 28, no. 2, pp. 345-350, 1989.
34. B. R. Rau, "Iterative modulo scheduling," Hewlett-Packard Res. Lab., Tech. Rep. HPL-94-115, 1995.
35. R. Schreiber and J. Dongarra, "Automatic blocking of nested loops," Univ. Tennessee, Knoxville, Tech. Rep. CS-90-108, 1990.
36. R. C. Whaley, private communication, 2004.
37. _, x86 optimizations, part 1. [Online], Available: http://sourceforge. net/mailarchive/forum.php?thread_id=1569256&forum_id=426
38. _, User contribution to ATLAS. [Online]. Available: http://math-atlas. sourceforge.net/devel/atlas_contrib
39. R. C. Whaley and A. Petitet, "Minimizing development and maintenance costs in supporting persistently optimized BLAS," Softw. Pract. Exper., to be published.
40. R. C. Whaley, A. Petitet, and J. J. Dongarra, "Automated empirical optimization of software and the ATLAS project," Parallel Comput, vol. 27, no. 1-2, pp. 3-35, 2001.
41. M. E. Wolf and M. S. Lam, "An algorithmic approach to compund loop transformations," in Advances in Languages and Compilers for Parallel Computing. Cambridge, MA: MIT Press, 1991, pp. 243-259.
42. M. Wolfe, "Iteration space tiling for memory hierarchies," in Proc. 3rd SIAM Conf. Parallel Processing for Scientific Computing, 1987, pp. 357-361.
43. K. Yotov, K. Pingali, and P. Stodghill, "X-Ray: A tool for automatic measurement of architectural parameters," Comput. Sci., Cornell Univ., Tech. Rep. TR2004-1966, 2004.