Method and structure for producing high performance linear...

Details Method and structure for producing high performance linear... Method and structure for producing high performance linear...

2003-09-29

2009-02-10

Do, Chat C (Department: 2193)

Electrical computers: arithmetic processing and calculating

Electrical digital calculating computer

Particular function performed

Reexamination Certificate

active

07490120

ABSTRACT:
A method (and structure) for improving at least one of speed and efficiency when executing level 3 dense linear algebra subroutines on a computer. An optimal matrix subroutine is selected from among a plurality of matrix subroutines stored in a memory that could alternatively perform a level 3 matrix multiplication or factorization processing.

REFERENCES:
patent: 5025407 (1991-06-01), Gulley et al.
patent: 5099447 (1992-03-01), Myszewski
patent: 5513366 (1996-04-01), Agarwal et al.
patent: 5644517 (1997-07-01), Ho
patent: 5825677 (1998-10-01), Agarwal et al.
patent: 5944819 (1999-08-01), Kumar et al.
patent: 5983230 (1999-11-01), Gilbert et al.
patent: 6021420 (2000-02-01), Takamuki
patent: 6115730 (2000-09-01), Dhablania et al.
patent: 6357041 (2002-03-01), Pingali et al.
patent: 6470368 (2002-10-01), Garg et al.
patent: 6507892 (2003-01-01), Mulla et al.
patent: 6601080 (2003-07-01), Garg
patent: 6675106 (2004-01-01), Keenan et al.
patent: 7028168 (2006-04-01), Wadleigh
patent: 7031994 (2006-04-01), Lao et al.
patent: 2003/0088600 (2003-05-01), Lao et al.
patent: 2004/0148324 (2004-07-01), Garg
Vinod et al., A framework for high-performance matrix multiplication based on hierarchical abstractions, algorithms and optimized low-level kernels, 2002, Concurrency and Computation: Practice and Experience 14(10): 805-839.
Phillip et al., PLAPACK: Parallel Linear Algebra Package Design Overview, 1997, IEEE Proceedings of the ACM/IEEE SC97 Conference.
Agarwal et al., A High Performance Algorithm Using Pre-Processing for the Sparse Matrix-Vector Multiplication, 1992, IEEE, pp. 32-41.
Volume I, “Foundations of Computer-Aided Chemical Process Design” Copyright 1981, American Institute of Chemical Engineering (2-Volume Series), pp. 77-143.
Juan J. Navarro, Elena Garcia, and Josep R. Herrero, “Data Prefetching and Multilevel Blocking for Linear Algebra Operations” in International Conference on Supercomputing (ICS'96), pp. 109-116, May 1996.
John A. Gunnels, Greg M. Henry, and Robert A. van de Geijn, “A Family of High-Performance Matrix Multiplication Algorithms,” ICCS 2001, LNCS 2073, pp. 51-60, 2001 (copyright Springer-Verlag Berlin Heidelberg 2001).
Fred G. Gustavson and Andre Henriksson and Isak Jonsson and Bo Kagstrom and Per Ling: Superscalar GEMM-based Level 3 BLAS The On-going Evolution of a Portable and High-Performance Library (1998); Applied Parallel Computing, Published 1998, Springer, pp. 207-215.
Gunnels, et al., “A Family of High-Performance Matrix Multiplication Algorithms”, ICCS 2001, LNCS 2073, pp. 51-60, 2001 (also available at http://www.cs.utexas.edu/users/flame/pubs/ICCS2001.pdf).
Gunnels, et al., “A Novel Theoretical Model Produces Matrix Multiplication Algorithms That Predict Current Practice”, IBM Research Report RC23443 (W0411-176), Nov. 19, 2004.
“Improving performance of linear algebra algorithms for dense matrices, using algorithmic prefetch” R. C. Agarwal, F. G. Gustavson, M. Zubair; IBM Journal of Research and Development; vol. 38, Issue 3 (May 1994); pp. 265-275; Year of Publication: 1994.

Affiliated with

Also associated with

Rating

Method and structure for producing high performance linear... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Method and structure for producing high performance linear..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and structure for producing high performance linear... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-4121379