<div class="csl-bib-body">
<div class="csl-entry">Hunold, S. (2023). Verifying Performance Guidelines for MPI Collectives at Scale. In <i>Proceedings of 2023 SC23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis (SC23 Workshops)</i> (pp. 1264–1268). ACM. https://doi.org/10.1145/3624062.3625532</div>
</div>
-
dc.identifier.uri
http://hdl.handle.net/20.500.12708/190663
-
dc.description.abstract
MPI collective communication operations are crucial for high-performance computing, making the efficient implementation of collective algorithms essential for optimal application performance. While most MPI libraries provide several algorithms for a specific collective operation, each may work better in a specific scenario. Therefore, selecting the most suitable algorithm for each use case is important. However, even the best algorithm in a given MPI library’s set may deliver suboptimal performance.
Self-consistent MPI performance guidelines are general expectations that collectives must meet to be deemed performance-consistent. Specifically, a specialized collective call should not be slower than its less specialized counterparts. This paper introduces a tool for assessing the performance consistency of MPI collectives in a statistically sound manner. Through a case study, we demonstrate the current state of MPI performance consistency for three TOP500 machines.
en
dc.description.sponsorship
FWF - Österr. Wissenschaftsfonds
-
dc.language.iso
en
-
dc.rights.uri
http://creativecommons.org/licenses/by/4.0/
-
dc.subject
MPI
en
dc.subject
collective communication
en
dc.subject
performance analysis
en
dc.subject
benchmarking
en
dc.subject
performance guidelines
en
dc.title
Verifying Performance Guidelines for MPI Collectives at Scale
en
dc.type
Inproceedings
en
dc.type
Konferenzbeitrag
de
dc.rights.license
Creative Commons Namensnennung 4.0 International
de
dc.rights.license
Creative Commons Attribution 4.0 International
en
dc.relation.isbn
9798400707858
-
dc.relation.doi
10.1145/3624062
-
dc.description.startpage
1264
-
dc.description.endpage
1268
-
dc.relation.grantno
P33884-N
-
dc.rights.holder
2023 Copyright held by the owner/author(s).
-
dc.type.category
Full-Paper Contribution
-
tuw.booktitle
Proceedings of 2023 SC23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis (SC23 Workshops)
-
tuw.peerreviewed
true
-
tuw.relation.publisher
ACM
-
tuw.relation.publisherplace
New York
-
tuw.project.title
Offline- und Online-Autotuning von Parallelen Programmen
-
tuw.researchTopic.id
I2
-
tuw.researchTopic.id
C5
-
tuw.researchTopic.name
Computer Engineering and Software-Intensive Systems
-
tuw.researchTopic.name
Computer Science Foundations
-
tuw.researchTopic.value
90
-
tuw.researchTopic.value
10
-
tuw.linking
https://doi.org/10.5281/zenodo.8384541
-
tuw.publication.orgunit
E191-04 - Forschungsbereich Parallel Computing
-
tuw.publisher.doi
10.1145/3624062.3625532
-
dc.description.numberOfPages
5
-
tuw.author.orcid
0000-0002-5280-3855
-
dc.rights.identifier
CC BY 4.0
de
dc.rights.identifier
CC BY 4.0
en
tuw.event.name
SC-W 2023: Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis