ABSTRACT
This paper describes the vectorization of the BLAS, a set of basic linear algebra subprograms for FORTRAN usage. The efficiency of the BLAS, as standard available on the CDC Cyber 205, has been examined, and suggestions for improvements were given. This examination has led to the vectorized BLAS as presented here. Moreover, this version admits negative increment values; that is, vectors can also be treated in reverse order. The number of data movements has been kept to a minimum. This BLAS version has been written in CDC FORTRAN 200. It has been optimized for a 1-pipe Cyber 205, but is also appropriate for the 2- and 4-pipe versions.
REFERENCES
Note: OCR errors may be found in this Reference List extracted from the full text article. ACM has opted to expose the complete List rather than only correct and linked references.
| |
1
|
AMERICAN NATIONAL STANDARDS INSTITUTE. American National Standard Programming Language FORTRAN, ANSI X3.g-1978. ANSI, New York, 1978.
|
| |
2
|
CDC. Cyber 200 ASSEMBLER reference manual. Ver. 2. Publ. 60485010, CDC, Sunnyvale, Calif., 1981.
|
| |
3
|
CDC. Cyber 200 FORTRAN reference manual. Ver. 1. Publ. 60480200B, CDC, Sunnyvale, Calif., 1983.
|
| |
4
|
CENTER FOR MATHEMATICS AND COMPUTER SCIENCE. A Library of NUMerical Software for VECtor and Parallel Computers in FORTRAN. Center for Mathematics and Computer Science, Amsterdam, 1985.
|
| |
5
|
DONGARRA, J. J., BUNCH, J. R., MOLER, C. B., AND STEWART, G.W. Linpack User's Guide. SIAM, Philadelphia, Pa., 1979.
|
| |
6
|
DONGARRA, J. J., Du CROZ, J., HAMMERLING, S., AND HANSON, R.J. An extended set of Fortran basic linear algebra subprograms. Tech. Memo. 41 (rev. 3), Argonne National Laboratory, Argonne, Ill., Nov. 1986.
|
 |
7
|
|
 |
8
|
|
| |
9
|
LOUTER-NOOL, M. Basic linear algebra subprograms (BLAS) on the CDC CYBER 205. Parallel Comput. 4 (1987), 143-165.
|
| |
10
|
SCHLICHTING, J. Double precision BLAS; Implementation design for double precision basic linear algebra subprograms on the Control Data Cyber 205. In Applications on Vector and Parallel Computers, H. J. J. te Riele, Th.J. Dekker, and H. A. van der Vorst, Eds., North-Holland, Amsterdam, 1987.
|
REVIEW
"A. Chris Rolls Newbery : Reviewer"
This paper examines the efficiency of the basic linear algebra
subprograms (BLAS) that are available on the CDC Cyber 205. By use of
vectorization the speed can be increased, in some cases by more than
fifty percent. New capabilities are also ad
more...
Peer to Peer - Readers of this Article have also read:
-
Data structures for quadtree approximation and compression
Communications of the ACM
28, 9
Hanan Samet
-
A hierarchical single-key-lock access control using the Chinese remainder theorem
Proceedings of the 1992 ACM/SIGAPP Symposium on Applied computing
Kim S. Lee
, Huizhu Lu
, D. D. Fisher
-
The GemStone object database management system
Communications of the ACM
34, 10
Paul Butterworth
, Allen Otis
, Jacob Stein
-
Putting innovation to work: adoption strategies for multimedia communication systems
Communications of the ACM
34, 12
Ellen Francik
, Susan Ehrlich Rudman
, Donna Cooper
, Stephen Levine
-
An intelligent component database for behavioral synthesis
Proceedings of the 27th ACM/IEEE Design Automation Conference on
Gwo-Dong Chen
, Daniel D. Gajski
|