Sökning: id:"swepub:oai:DiVA.org:kth-338614" >
Exploring Technique...
Exploring Techniques for the Analysis of Spontaneous Asynchronicity in MPI-Parallel Applications
-
- Afzal, Ayesha (författare)
- Erlangen National High Performance Computing Center (NHR@FAU), 91058, Erlangen, Germany
-
- Hager, Georg (författare)
- Erlangen National High Performance Computing Center (NHR@FAU), 91058, Erlangen, Germany
-
- Wellein, Gerhard (författare)
- Erlangen National High Performance Computing Center (NHR@FAU), 91058, Erlangen, Germany; Department of Computer Science, University of Erlangen-Nürnberg, 91058, Erlangen, Germany
-
visa fler...
-
- Markidis, Stefano (författare)
- KTH,Beräkningsvetenskap och beräkningsteknik (CST)
-
visa färre...
-
(creator_code:org_t)
- Springer Nature, 2023
- 2023
- Engelska.
-
Ingår i: Parallel Processing and Applied Mathematics - 14th International Conference, PPAM 2022, Revised Selected Papers. - : Springer Nature. ; , s. 155-170
- Relaterad länk:
-
https://doi.org/10.1...
-
visa fler...
-
https://urn.kb.se/re...
-
https://doi.org/10.1...
-
visa färre...
Abstract
Ämnesord
Stäng
- This paper studies the utility of using data analytics and machine learning techniques for identifying, classifying, and characterizing the dynamics of large-scale parallel (MPI) programs. To this end, we run microbenchmarks and realistic proxy applications with the regular compute-communicate structure on two different supercomputing platforms and choose the per-process performance and MPI time per time step as relevant observables. Using principal component analysis, clustering techniques, correlation functions, and a new “phase space plot,” we show how desynchronization patterns (or lack thereof) can be readily identified from a data set that is much smaller than a full MPI trace. Our methods also lead the way towards a more general classification of parallel program dynamics.
Ämnesord
- NATURVETENSKAP -- Data- och informationsvetenskap -- Datavetenskap (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Computer Sciences (hsv//eng)
- NATURVETENSKAP -- Data- och informationsvetenskap -- Datorteknik (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Computer Engineering (hsv//eng)
Nyckelord
- Asynchronous MPI execution
- Data analytic techniques
- Machine learning techniques
- Parallel distributed computing
- Scalability and bottleneck
Publikations- och innehållstyp
- ref (ämneskategori)
- kon (ämneskategori)