Fast sparse matrix factorization on modern workstations:
The performance of workstation-class machines has experienced a dramatic increase in the recent past. Relatively inexpensive machines which offer 14 MIPS and 2 MFLOPS performance are now available, and machines with even higher performance are not far off. One important characteristic of these machi...
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Buch |
| Sprache: | English |
| Veröffentlicht: |
Stanford, Calif.
1989
|
| Schriftenreihe: | Stanford University / Computer Science Department: Report STAN-CS
1286 |
| Schlagworte: | |
| Zusammenfassung: | The performance of workstation-class machines has experienced a dramatic increase in the recent past. Relatively inexpensive machines which offer 14 MIPS and 2 MFLOPS performance are now available, and machines with even higher performance are not far off. One important characteristic of these machines is that they rely on a small amount of high-speed cache memory for their high performance. In this paper, we consider the problem of Cholesky factorization of a large sparse positive definite system of equations on a high performance workstation. We find that the major factor limiting performance is the cost of moving data between memory and the processor. We use two techniques to address this limitation; we decrease the number of memory references and we improve cache behavior to decrease the cost of each reference. When run on benchmarks from the Harwell-Boeing Sparse Matrix Collection, the resulting factorization code is almost three times as fast as SPARSPAK on a DECStation 3100. We believe that the issues brought up in this paper will play an important role in the effective use of high performance workstations on large numerical problems. |
| Beschreibung: | 15 S. |
Internformat
MARC
| LEADER | 00000nam a2200000 cb4500 | ||
|---|---|---|---|
| 001 | BV008948805 | ||
| 003 | DE-604 | ||
| 005 | 00000000000000.0 | ||
| 007 | t | ||
| 008 | 940206s1989 |||| 00||| eng d | ||
| 035 | |a (OCoLC)21236170 | ||
| 035 | |a (DE-599)BVBBV008948805 | ||
| 040 | |a DE-604 |b ger |e rakddb | ||
| 041 | 0 | |a eng | |
| 049 | |a DE-29T | ||
| 100 | 1 | |a Rothberg, Edward |e Verfasser |4 aut | |
| 245 | 1 | 0 | |a Fast sparse matrix factorization on modern workstations |c Edward Rothberg and Anoop Gupta |
| 264 | 1 | |a Stanford, Calif. |c 1989 | |
| 300 | |a 15 S. | ||
| 336 | |b txt |2 rdacontent | ||
| 337 | |b n |2 rdamedia | ||
| 338 | |b nc |2 rdacarrier | ||
| 490 | 1 | |a Stanford University / Computer Science Department: Report STAN-CS |v 1286 | |
| 520 | 3 | |a The performance of workstation-class machines has experienced a dramatic increase in the recent past. Relatively inexpensive machines which offer 14 MIPS and 2 MFLOPS performance are now available, and machines with even higher performance are not far off. One important characteristic of these machines is that they rely on a small amount of high-speed cache memory for their high performance. In this paper, we consider the problem of Cholesky factorization of a large sparse positive definite system of equations on a high performance workstation. We find that the major factor limiting performance is the cost of moving data between memory and the processor. We use two techniques to address this limitation; we decrease the number of memory references and we improve cache behavior to decrease the cost of each reference. When run on benchmarks from the Harwell-Boeing Sparse Matrix Collection, the resulting factorization code is almost three times as fast as SPARSPAK on a DECStation 3100. We believe that the issues brought up in this paper will play an important role in the effective use of high performance workstations on large numerical problems. | |
| 650 | 4 | |a Microcomputer workstations | |
| 650 | 4 | |a Sparse matrices | |
| 700 | 1 | |a Gupta, Anoop |e Verfasser |4 aut | |
| 810 | 2 | |a Computer Science Department: Report STAN-CS |t Stanford University |v 1286 |w (DE-604)BV008928280 |9 1286 | |
| 999 | |a oai:aleph.bib-bvb.de:BVB01-005904533 | ||
Datensatz im Suchindex
| _version_ | 1804123281683382272 |
|---|---|
| any_adam_object | |
| author | Rothberg, Edward Gupta, Anoop |
| author_facet | Rothberg, Edward Gupta, Anoop |
| author_role | aut aut |
| author_sort | Rothberg, Edward |
| author_variant | e r er a g ag |
| building | Verbundindex |
| bvnumber | BV008948805 |
| ctrlnum | (OCoLC)21236170 (DE-599)BVBBV008948805 |
| format | Book |
| fullrecord | <?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>02172nam a2200313 cb4500</leader><controlfield tag="001">BV008948805</controlfield><controlfield tag="003">DE-604</controlfield><controlfield tag="005">00000000000000.0</controlfield><controlfield tag="007">t</controlfield><controlfield tag="008">940206s1989 |||| 00||| eng d</controlfield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(OCoLC)21236170</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)BVBBV008948805</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-604</subfield><subfield code="b">ger</subfield><subfield code="e">rakddb</subfield></datafield><datafield tag="041" ind1="0" ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="049" ind1=" " ind2=" "><subfield code="a">DE-29T</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Rothberg, Edward</subfield><subfield code="e">Verfasser</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Fast sparse matrix factorization on modern workstations</subfield><subfield code="c">Edward Rothberg and Anoop Gupta</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="a">Stanford, Calif.</subfield><subfield code="c">1989</subfield></datafield><datafield tag="300" ind1=" " ind2=" "><subfield code="a">15 S.</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="490" ind1="1" ind2=" "><subfield code="a">Stanford University / Computer Science Department: Report STAN-CS</subfield><subfield code="v">1286</subfield></datafield><datafield tag="520" ind1="3" ind2=" "><subfield code="a">The performance of workstation-class machines has experienced a dramatic increase in the recent past. Relatively inexpensive machines which offer 14 MIPS and 2 MFLOPS performance are now available, and machines with even higher performance are not far off. One important characteristic of these machines is that they rely on a small amount of high-speed cache memory for their high performance. In this paper, we consider the problem of Cholesky factorization of a large sparse positive definite system of equations on a high performance workstation. We find that the major factor limiting performance is the cost of moving data between memory and the processor. We use two techniques to address this limitation; we decrease the number of memory references and we improve cache behavior to decrease the cost of each reference. When run on benchmarks from the Harwell-Boeing Sparse Matrix Collection, the resulting factorization code is almost three times as fast as SPARSPAK on a DECStation 3100. We believe that the issues brought up in this paper will play an important role in the effective use of high performance workstations on large numerical problems.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Microcomputer workstations</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Sparse matrices</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Gupta, Anoop</subfield><subfield code="e">Verfasser</subfield><subfield code="4">aut</subfield></datafield><datafield tag="810" ind1="2" ind2=" "><subfield code="a">Computer Science Department: Report STAN-CS</subfield><subfield code="t">Stanford University</subfield><subfield code="v">1286</subfield><subfield code="w">(DE-604)BV008928280</subfield><subfield code="9">1286</subfield></datafield><datafield tag="999" ind1=" " ind2=" "><subfield code="a">oai:aleph.bib-bvb.de:BVB01-005904533</subfield></datafield></record></collection> |
| id | DE-604.BV008948805 |
| illustrated | Not Illustrated |
| indexdate | 2024-07-09T17:27:17Z |
| institution | BVB |
| language | English |
| oai_aleph_id | oai:aleph.bib-bvb.de:BVB01-005904533 |
| oclc_num | 21236170 |
| open_access_boolean | |
| owner | DE-29T |
| owner_facet | DE-29T |
| physical | 15 S. |
| publishDate | 1989 |
| publishDateSearch | 1989 |
| publishDateSort | 1989 |
| record_format | marc |
| series2 | Stanford University / Computer Science Department: Report STAN-CS |
| spelling | Rothberg, Edward Verfasser aut Fast sparse matrix factorization on modern workstations Edward Rothberg and Anoop Gupta Stanford, Calif. 1989 15 S. txt rdacontent n rdamedia nc rdacarrier Stanford University / Computer Science Department: Report STAN-CS 1286 The performance of workstation-class machines has experienced a dramatic increase in the recent past. Relatively inexpensive machines which offer 14 MIPS and 2 MFLOPS performance are now available, and machines with even higher performance are not far off. One important characteristic of these machines is that they rely on a small amount of high-speed cache memory for their high performance. In this paper, we consider the problem of Cholesky factorization of a large sparse positive definite system of equations on a high performance workstation. We find that the major factor limiting performance is the cost of moving data between memory and the processor. We use two techniques to address this limitation; we decrease the number of memory references and we improve cache behavior to decrease the cost of each reference. When run on benchmarks from the Harwell-Boeing Sparse Matrix Collection, the resulting factorization code is almost three times as fast as SPARSPAK on a DECStation 3100. We believe that the issues brought up in this paper will play an important role in the effective use of high performance workstations on large numerical problems. Microcomputer workstations Sparse matrices Gupta, Anoop Verfasser aut Computer Science Department: Report STAN-CS Stanford University 1286 (DE-604)BV008928280 1286 |
| spellingShingle | Rothberg, Edward Gupta, Anoop Fast sparse matrix factorization on modern workstations Microcomputer workstations Sparse matrices |
| title | Fast sparse matrix factorization on modern workstations |
| title_auth | Fast sparse matrix factorization on modern workstations |
| title_exact_search | Fast sparse matrix factorization on modern workstations |
| title_full | Fast sparse matrix factorization on modern workstations Edward Rothberg and Anoop Gupta |
| title_fullStr | Fast sparse matrix factorization on modern workstations Edward Rothberg and Anoop Gupta |
| title_full_unstemmed | Fast sparse matrix factorization on modern workstations Edward Rothberg and Anoop Gupta |
| title_short | Fast sparse matrix factorization on modern workstations |
| title_sort | fast sparse matrix factorization on modern workstations |
| topic | Microcomputer workstations Sparse matrices |
| topic_facet | Microcomputer workstations Sparse matrices |
| volume_link | (DE-604)BV008928280 |
| work_keys_str_mv | AT rothbergedward fastsparsematrixfactorizationonmodernworkstations AT guptaanoop fastsparsematrixfactorizationonmodernworkstations |