Efficient doacross execution on distributed shared-memory multiprocessors:
Abstract: "Doacross can exploit the parallelism of a Do loop with cross-iteration dependences. All previous approaches to Doacross execution assumed the uniform memory access (UMA) architecture. In this paper, we propose a Doacross execution scheme for the distributed shared memory (DSM) archit...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Book |
| Language: | English |
| Published: |
Urbana, Ill.
1991
|
| Series: | Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report
1072 |
| Subjects: | |
| Summary: | Abstract: "Doacross can exploit the parallelism of a Do loop with cross-iteration dependences. All previous approaches to Doacross execution assumed the uniform memory access (UMA) architecture. In this paper, we propose a Doacross execution scheme for the distributed shared memory (DSM) architecture. Our scheme uses direct communication and static message passing. Our communication is direct because a producer writes (or sends) data directly to the nearest shared memory (NeSM) of its consumer. Our message passing is static in the sense that a message is bound to a buffer location determined at compile time We present compiler algorithms that transform a Doacross loop into the proposed execution scheme and derive conditions that can reduce the extra storage needed for message buffers. The preliminary results show that the proposed scheme can outperform approaches that ignore the features of the DSM architecture, i.e. simply utilize DSM architecture as UMA architecture. |
| Physical Description: | ca. 15 Bl. |
Staff View
MARC
| LEADER | 00000nam a2200000 cb4500 | ||
|---|---|---|---|
| 001 | BV008992785 | ||
| 003 | DE-604 | ||
| 005 | 00000000000000.0 | ||
| 007 | t | ||
| 008 | 940206s1991 |||| 00||| eng d | ||
| 035 | |a (OCoLC)25761811 | ||
| 035 | |a (DE-599)BVBBV008992785 | ||
| 040 | |a DE-604 |b ger |e rakddb | ||
| 041 | 0 | |a eng | |
| 049 | |a DE-29T | ||
| 100 | 1 | |a Su, Hong-Men |e Verfasser |4 aut | |
| 245 | 1 | 0 | |a Efficient doacross execution on distributed shared-memory multiprocessors |c Hong-Men Su and Pen-Chung Yew |
| 264 | 1 | |a Urbana, Ill. |c 1991 | |
| 300 | |a ca. 15 Bl. | ||
| 336 | |b txt |2 rdacontent | ||
| 337 | |b n |2 rdamedia | ||
| 338 | |b nc |2 rdacarrier | ||
| 490 | 1 | |a Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report |v 1072 | |
| 520 | 3 | |a Abstract: "Doacross can exploit the parallelism of a Do loop with cross-iteration dependences. All previous approaches to Doacross execution assumed the uniform memory access (UMA) architecture. In this paper, we propose a Doacross execution scheme for the distributed shared memory (DSM) architecture. Our scheme uses direct communication and static message passing. Our communication is direct because a producer writes (or sends) data directly to the nearest shared memory (NeSM) of its consumer. Our message passing is static in the sense that a message is bound to a buffer location determined at compile time | |
| 520 | 3 | |a We present compiler algorithms that transform a Doacross loop into the proposed execution scheme and derive conditions that can reduce the extra storage needed for message buffers. The preliminary results show that the proposed scheme can outperform approaches that ignore the features of the DSM architecture, i.e. simply utilize DSM architecture as UMA architecture. | |
| 650 | 4 | |a Datenverarbeitung | |
| 650 | 4 | |a Electronic data processing |x Distributed processing | |
| 650 | 4 | |a Multiprocessors | |
| 700 | 1 | |a Yew, Pen-Chung |e Verfasser |4 aut | |
| 830 | 0 | |a Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report |v 1072 |w (DE-604)BV008930033 |9 1072 | |
| 999 | |a oai:aleph.bib-bvb.de:BVB01-005941704 | ||
Record in the Search Index
| _version_ | 1804123335299170304 |
|---|---|
| any_adam_object | |
| author | Su, Hong-Men Yew, Pen-Chung |
| author_facet | Su, Hong-Men Yew, Pen-Chung |
| author_role | aut aut |
| author_sort | Su, Hong-Men |
| author_variant | h m s hms p c y pcy |
| building | Verbundindex |
| bvnumber | BV008992785 |
| ctrlnum | (OCoLC)25761811 (DE-599)BVBBV008992785 |
| format | Book |
| fullrecord | <?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>02112nam a2200337 cb4500</leader><controlfield tag="001">BV008992785</controlfield><controlfield tag="003">DE-604</controlfield><controlfield tag="005">00000000000000.0</controlfield><controlfield tag="007">t</controlfield><controlfield tag="008">940206s1991 |||| 00||| eng d</controlfield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(OCoLC)25761811</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)BVBBV008992785</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">Su, Hong-Men</subfield><subfield code="e">Verfasser</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Efficient doacross execution on distributed shared-memory multiprocessors</subfield><subfield code="c">Hong-Men Su and Pen-Chung Yew</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="a">Urbana, Ill.</subfield><subfield code="c">1991</subfield></datafield><datafield tag="300" ind1=" " ind2=" "><subfield code="a">ca. 15 Bl.</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">Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report</subfield><subfield code="v">1072</subfield></datafield><datafield tag="520" ind1="3" ind2=" "><subfield code="a">Abstract: "Doacross can exploit the parallelism of a Do loop with cross-iteration dependences. All previous approaches to Doacross execution assumed the uniform memory access (UMA) architecture. In this paper, we propose a Doacross execution scheme for the distributed shared memory (DSM) architecture. Our scheme uses direct communication and static message passing. Our communication is direct because a producer writes (or sends) data directly to the nearest shared memory (NeSM) of its consumer. Our message passing is static in the sense that a message is bound to a buffer location determined at compile time</subfield></datafield><datafield tag="520" ind1="3" ind2=" "><subfield code="a">We present compiler algorithms that transform a Doacross loop into the proposed execution scheme and derive conditions that can reduce the extra storage needed for message buffers. The preliminary results show that the proposed scheme can outperform approaches that ignore the features of the DSM architecture, i.e. simply utilize DSM architecture as UMA architecture.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Datenverarbeitung</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Electronic data processing</subfield><subfield code="x">Distributed processing</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Multiprocessors</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Yew, Pen-Chung</subfield><subfield code="e">Verfasser</subfield><subfield code="4">aut</subfield></datafield><datafield tag="830" ind1=" " ind2="0"><subfield code="a">Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report</subfield><subfield code="v">1072</subfield><subfield code="w">(DE-604)BV008930033</subfield><subfield code="9">1072</subfield></datafield><datafield tag="999" ind1=" " ind2=" "><subfield code="a">oai:aleph.bib-bvb.de:BVB01-005941704</subfield></datafield></record></collection> |
| id | DE-604.BV008992785 |
| illustrated | Not Illustrated |
| indexdate | 2024-07-09T17:28:08Z |
| institution | BVB |
| language | English |
| oai_aleph_id | oai:aleph.bib-bvb.de:BVB01-005941704 |
| oclc_num | 25761811 |
| open_access_boolean | |
| owner | DE-29T |
| owner_facet | DE-29T |
| physical | ca. 15 Bl. |
| publishDate | 1991 |
| publishDateSearch | 1991 |
| publishDateSort | 1991 |
| record_format | marc |
| series | Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report |
| series2 | Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report |
| spelling | Su, Hong-Men Verfasser aut Efficient doacross execution on distributed shared-memory multiprocessors Hong-Men Su and Pen-Chung Yew Urbana, Ill. 1991 ca. 15 Bl. txt rdacontent n rdamedia nc rdacarrier Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report 1072 Abstract: "Doacross can exploit the parallelism of a Do loop with cross-iteration dependences. All previous approaches to Doacross execution assumed the uniform memory access (UMA) architecture. In this paper, we propose a Doacross execution scheme for the distributed shared memory (DSM) architecture. Our scheme uses direct communication and static message passing. Our communication is direct because a producer writes (or sends) data directly to the nearest shared memory (NeSM) of its consumer. Our message passing is static in the sense that a message is bound to a buffer location determined at compile time We present compiler algorithms that transform a Doacross loop into the proposed execution scheme and derive conditions that can reduce the extra storage needed for message buffers. The preliminary results show that the proposed scheme can outperform approaches that ignore the features of the DSM architecture, i.e. simply utilize DSM architecture as UMA architecture. Datenverarbeitung Electronic data processing Distributed processing Multiprocessors Yew, Pen-Chung Verfasser aut Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report 1072 (DE-604)BV008930033 1072 |
| spellingShingle | Su, Hong-Men Yew, Pen-Chung Efficient doacross execution on distributed shared-memory multiprocessors Center for Supercomputing Research and Development <Urbana, Ill,>: CSRD report Datenverarbeitung Electronic data processing Distributed processing Multiprocessors |
| title | Efficient doacross execution on distributed shared-memory multiprocessors |
| title_auth | Efficient doacross execution on distributed shared-memory multiprocessors |
| title_exact_search | Efficient doacross execution on distributed shared-memory multiprocessors |
| title_full | Efficient doacross execution on distributed shared-memory multiprocessors Hong-Men Su and Pen-Chung Yew |
| title_fullStr | Efficient doacross execution on distributed shared-memory multiprocessors Hong-Men Su and Pen-Chung Yew |
| title_full_unstemmed | Efficient doacross execution on distributed shared-memory multiprocessors Hong-Men Su and Pen-Chung Yew |
| title_short | Efficient doacross execution on distributed shared-memory multiprocessors |
| title_sort | efficient doacross execution on distributed shared memory multiprocessors |
| topic | Datenverarbeitung Electronic data processing Distributed processing Multiprocessors |
| topic_facet | Datenverarbeitung Electronic data processing Distributed processing Multiprocessors |
| volume_link | (DE-604)BV008930033 |
| work_keys_str_mv | AT suhongmen efficientdoacrossexecutionondistributedsharedmemorymultiprocessors AT yewpenchung efficientdoacrossexecutionondistributedsharedmemorymultiprocessors |