MPI-2 One-Sided Usage and Implementation for Read Modify Write Operations: A Case Study with HPCC
MPI-2’s One-sided communication interface is being explored in scientific applications. One of the important operations in a one sided model is read-modify-write. MPI-2 semantics provide MPI_Put, MPI_Get and MPI_Accumulate operations which can be used to implement read-modify-write functionality. Th...
Saved in:
Published in | Recent Advances in Parallel Virtual Machine and Message Passing Interface pp. 251 - 259 |
---|---|
Main Authors | , , , |
Format | Book Chapter |
Language | English |
Published |
Berlin, Heidelberg
Springer Berlin Heidelberg
|
Series | Lecture Notes in Computer Science |
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | MPI-2’s One-sided communication interface is being explored in scientific applications. One of the important operations in a one sided model is read-modify-write. MPI-2 semantics provide MPI_Put, MPI_Get and MPI_Accumulate operations which can be used to implement read-modify-write functionality. The different strategies yield varying performance benefits depending on the underlying one-sided implementation. We use HPCC Random Access benchmark which primarily uses read-modify-write operations as a case study for evaluating the different implementation strategies in this paper. Currently this benchmark is implemented based on MPI two-sided semantics. In this work we design and evaluate MPI-2 versions of the HPCC Random Access benchmark using one-sided operations. To improve the performance, we explore two different optimizations: (i) software based aggregation and (ii) hardware-based atomic operations. We evaluate our different approaches on an InfiniBand cluster. The software based aggregation outperforms the basic one sided scheme without aggregation by a factor of 4.38. The hardware based scheme shows an improvement by a factor of 2.62 as compared to the basic one sided scheme. |
---|---|
AbstractList | MPI-2’s One-sided communication interface is being explored in scientific applications. One of the important operations in a one sided model is read-modify-write. MPI-2 semantics provide MPI_Put, MPI_Get and MPI_Accumulate operations which can be used to implement read-modify-write functionality. The different strategies yield varying performance benefits depending on the underlying one-sided implementation. We use HPCC Random Access benchmark which primarily uses read-modify-write operations as a case study for evaluating the different implementation strategies in this paper. Currently this benchmark is implemented based on MPI two-sided semantics. In this work we design and evaluate MPI-2 versions of the HPCC Random Access benchmark using one-sided operations. To improve the performance, we explore two different optimizations: (i) software based aggregation and (ii) hardware-based atomic operations. We evaluate our different approaches on an InfiniBand cluster. The software based aggregation outperforms the basic one sided scheme without aggregation by a factor of 4.38. The hardware based scheme shows an improvement by a factor of 2.62 as compared to the basic one sided scheme. |
Author | Narravula, Sundeep Santhanaraman, Gopalakrishnan Mamidala, Amith. R. Panda, Dhabaleswar K. |
Author_xml | – sequence: 1 givenname: Gopalakrishnan surname: Santhanaraman fullname: Santhanaraman, Gopalakrishnan email: santhana@cse.ohio-state.edu organization: Department of Computer Science and Engineering, The Ohio State University, Columbus, Ohio 43210, – sequence: 2 givenname: Sundeep surname: Narravula fullname: Narravula, Sundeep email: narravul@cse.ohio-state.edu organization: Department of Computer Science and Engineering, The Ohio State University, Columbus, Ohio 43210, – sequence: 3 givenname: Amith. R. surname: Mamidala fullname: Mamidala, Amith. R. email: mamidala@cse.ohio-state.edu organization: Department of Computer Science and Engineering, The Ohio State University, Columbus, Ohio 43210, – sequence: 4 givenname: Dhabaleswar K. surname: Panda fullname: Panda, Dhabaleswar K. email: panda@cse.ohio-state.edu organization: Department of Computer Science and Engineering, The Ohio State University, Columbus, Ohio 43210, |
BookMark | eNo1kMFOAjEQhqtiIiBv4KEvUG237bb1RjYiJBCISDw2hU5xFbpku8bw9i6oc5nk-yeTma-HOrGKgNAdo_eMUvVglCacSEGJkoLlxFieX6Aeb8kZiEvUZTljhHNhrtCgnf_PZNZBXcppRowS_AYNUvqgbXFmVC67yM0WE5LheQSyLD14vEpuC9hFjyf7ww72EBvXlFXEoarxCziPZ5UvwxG_1WUDeH6A-pynRzzEhUuAl82XP-LvsnnH40VR3KLr4HYJBn-9j1ajp9diTKbz50kxnJLEjG5IkK69OmhJ_ToYxYLwUmrHQZ-e0hsPGx6cAc-cU5kDI4LPNga4billmvdR9rs3HeoybqG266r6TJZRe5JoWymW21aLPTuzJ4n8ByhMYRg |
ContentType | Book Chapter |
Copyright | Springer-Verlag Berlin Heidelberg 2007 |
Copyright_xml | – notice: Springer-Verlag Berlin Heidelberg 2007 |
DOI | 10.1007/978-3-540-75416-9_36 |
DatabaseTitleList | |
DeliveryMethod | fulltext_linktorsrc |
Discipline | Computer Science |
EISBN | 3540754164 9783540754169 |
EISSN | 1611-3349 |
Editor | Cappello, Franck Dongarra, Jack Herault, Thomas |
Editor_xml | – sequence: 1 givenname: Franck surname: Cappello fullname: Cappello, Franck email: fci@lri.fr – sequence: 2 givenname: Thomas surname: Herault fullname: Herault, Thomas email: thomas.herault@lri.fr – sequence: 3 givenname: Jack surname: Dongarra fullname: Dongarra, Jack email: dongarra@cs.utk.edu |
EndPage | 259 |
GroupedDBID | -DT -GH -~X 1SB 29L 2HA 2HV 5QI 875 AASHB ABMNI ACGFS ADCXD AEFIE ALMA_UNASSIGNED_HOLDINGS EJD F5P FEDTE HVGLF LAS LDH P2P RIG RNI RSU SVGTG VI1 ~02 |
ID | FETCH-LOGICAL-s198t-f5a783f850dbf971f4d558a3e833498cdec3fa9ed1aa72ae94fd2c9e38fa90183 |
ISBN | 9783540754152 3540754156 |
ISSN | 0302-9743 |
IngestDate | Tue Oct 01 19:22:26 EDT 2024 |
IsPeerReviewed | true |
IsScholarly | true |
Language | English |
LinkModel | OpenURL |
MergedId | FETCHMERGED-LOGICAL-s198t-f5a783f850dbf971f4d558a3e833498cdec3fa9ed1aa72ae94fd2c9e38fa90183 |
Notes | This research is supported in part by Department of Energy’s grant #DE-FC02-06ER25749 and #DE-FC02-06ER25755; National Science Foundation’s grants #CNS-0403342 and #CCF-0702675; grants from Intel, Sun Microsystems, Cisco Systems, and Linux Networx; Equipment donations from Intel, Mellanox, AMD, Apple, IBM, Microway, PathScale, Silverstorm and Sun Microsystems. |
PageCount | 9 |
ParticipantIDs | springer_books_10_1007_978_3_540_75416_9_36 |
PublicationPlace | Berlin, Heidelberg |
PublicationPlace_xml | – name: Berlin, Heidelberg |
PublicationSeriesTitle | Lecture Notes in Computer Science |
PublicationSubtitle | 14th European PVM/MPI User’s Group Meeting, Paris, France, September 30 - October 3, 2007. Proceedings |
PublicationTitle | Recent Advances in Parallel Virtual Machine and Message Passing Interface |
Publisher | Springer Berlin Heidelberg |
Publisher_xml | – name: Springer Berlin Heidelberg |
RelatedPersons | Kleinberg, Jon M. Mattern, Friedemann Nierstrasz, Oscar Steffen, Bernhard Kittler, Josef Vardi, Moshe Y. Weikum, Gerhard Sudan, Madhu Naor, Moni Mitchell, John C. Terzopoulos, Demetri Pandu Rangan, C. Kanade, Takeo Hutchison, David Tygar, Doug |
RelatedPersons_xml | – sequence: 1 givenname: David surname: Hutchison fullname: Hutchison, David – sequence: 2 givenname: Takeo surname: Kanade fullname: Kanade, Takeo – sequence: 3 givenname: Josef surname: Kittler fullname: Kittler, Josef – sequence: 4 givenname: Jon M. surname: Kleinberg fullname: Kleinberg, Jon M. – sequence: 5 givenname: Friedemann surname: Mattern fullname: Mattern, Friedemann – sequence: 6 givenname: John C. surname: Mitchell fullname: Mitchell, John C. – sequence: 7 givenname: Moni surname: Naor fullname: Naor, Moni – sequence: 8 givenname: Oscar surname: Nierstrasz fullname: Nierstrasz, Oscar – sequence: 9 givenname: C. surname: Pandu Rangan fullname: Pandu Rangan, C. – sequence: 10 givenname: Bernhard surname: Steffen fullname: Steffen, Bernhard – sequence: 11 givenname: Madhu surname: Sudan fullname: Sudan, Madhu – sequence: 12 givenname: Demetri surname: Terzopoulos fullname: Terzopoulos, Demetri – sequence: 13 givenname: Doug surname: Tygar fullname: Tygar, Doug – sequence: 14 givenname: Moshe Y. surname: Vardi fullname: Vardi, Moshe Y. – sequence: 15 givenname: Gerhard surname: Weikum fullname: Weikum, Gerhard |
SSID | ssj0000319765 ssj0002792 |
Score | 1.6755117 |
Snippet | MPI-2’s One-sided communication interface is being explored in scientific applications. One of the important operations in a one sided model is... |
SourceID | springer |
SourceType | Publisher |
StartPage | 251 |
SubjectTerms | Atomic Operation Communication Operation Remote Memory Remote Node Software Aggregation |
Title | MPI-2 One-Sided Usage and Implementation for Read Modify Write Operations: A Case Study with HPCC |
URI | http://link.springer.com/10.1007/978-3-540-75416-9_36 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwnV1Lb9NAEF6lRVw48BZFBe2Bm-XIb3uReoiqQihNGtGm9GatvWs1InUr2wGV38qPYWZ3badthFQuVrKO7PV8zs5jv5kh5IPjSBEWQWwXXiLswPUymzlhbrMkz7gArzmPMTl5Mo3G8-DwPDwfDP6ssZZWTTbMf2_MK_kfVGEMcMUs2Qcg210UBuAz4AtHQBiOd4zf22FWUyIWiZXWSG_iK1rrjFfYGmVpnS0qlRcyUVRJw6bAZidI0AFzWdMEsEM175E9ASFf8BKuYaKin8GhXvIf2IO-7F-iKa8q_nO15JrUgzlo131c-3IhuD41wpjN0Po27HepwP1Xa9wFz0Av1b94ZX0drr-0k9kX27OOS2mfLASYwnM1Y0VZvmyJ7h05Evn_2MttUdxY3yuQjnV8LfULXZuEe1DRiil5o-PN49m-DgojOrLeOzIbKNOrRguw7XHRLnn34p_WP8qDmfBWHKK1srbA-qANwJ_SC6zUCiDCso6-LqPaLuqmJK4039hG1XOXbYJ3i2yW-tEW2YoZrLiPRgeHR2ddBBDzx-Kod9awlKPe89KzMplIatbR4y5Ip59iLQt00y3v7esrc-n0GXmCKTQUc1tAmM_JQJYvyNNWuNQI9yXhCm_a4U0V3hTwprfxpoA3RbypxpsqvGmP90c6oog2VWhTRJsi2q_I_NPB6f7YNj0_7NplSWMXIYfHLJLQEVnBYrcIRBgm3JcJgpLkQuZ-wZkULuexxyULCuHlTPoJjDqgn16T7fKqlG8ITUIROK7MmYhEkIBnHonQzSIOCINTUDg7xGpllOK_uE7bEt4g0dRPQaKpkmiKEn37oF_vku2mWsl3YLA22XuD_F8IqpCS |
link.rule.ids | 785,786,790,799,27956 |
linkProvider | Library Specific Holdings |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=bookitem&rft.title=Recent+Advances+in+Parallel+Virtual+Machine+and+Message+Passing+Interface&rft.au=Santhanaraman%2C+Gopalakrishnan&rft.au=Narravula%2C+Sundeep&rft.au=Mamidala%2C+Amith.+R.&rft.au=Panda%2C+Dhabaleswar+K.&rft.atitle=MPI-2+One-Sided+Usage+and+Implementation+for+Read+Modify+Write+Operations%3A+A+Case+Study+with+HPCC&rft.series=Lecture+Notes+in+Computer+Science&rft.pub=Springer+Berlin+Heidelberg&rft.isbn=9783540754152&rft.issn=0302-9743&rft.eissn=1611-3349&rft.spage=251&rft.epage=259&rft_id=info:doi/10.1007%2F978-3-540-75416-9_36 |
thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0302-9743&client=summon |
thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0302-9743&client=summon |
thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0302-9743&client=summon |