Energy-efficient multigrid smoothers and grid transfer operators on multi-core and GPU clusters
We investigate time and energy to solution for the CPU- and GPU-based execution of the compute intensive smoother and grid transfer operators in a geometric multigrid linear solver. We use a hybrid parallel implementation for both shared and distributed memory multi-core host systems comprising CUDA...
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Dokumenttyp: | Article (Journal) |
| Sprache: | Englisch |
| Veröffentlicht: |
2017
|
| In: |
Journal of parallel and distributed computing
Year: 2016, Jahrgang: 100, Pages: 181-192 |
| DOI: | 10.1016/j.jpdc.2016.05.006 |
| Online-Zugang: | Verlag, Volltext: http://dx.doi.org/10.1016/j.jpdc.2016.05.006 Verlag, Volltext: http://www.sciencedirect.com/science/article/pii/S0743731516300363 
|
| Verfasserangaben: | Martin Wlotzka, Vincent Heuveline |
MARC
| LEADER | 00000caa a2200000 c 4500 | ||
|---|---|---|---|
| 001 | 1581442211 | ||
| 003 | DE-627 | ||
| 005 | 20220815022625.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 181001r20172016xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1016/j.jpdc.2016.05.006 |2 doi | |
| 035 | |a (DE-627)1581442211 | ||
| 035 | |a (DE-576)511442211 | ||
| 035 | |a (DE-599)BSZ511442211 | ||
| 035 | |a (OCoLC)1341019456 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rda | ||
| 041 | |a eng | ||
| 084 | |a 29 |2 sdnb | ||
| 100 | 1 | |a Wlotzka, Martin |e VerfasserIn |0 (DE-588)1031120645 |0 (DE-627)735676135 |0 (DE-576)378546856 |4 aut | |
| 245 | 1 | 0 | |a Energy-efficient multigrid smoothers and grid transfer operators on multi-core and GPU clusters |c Martin Wlotzka, Vincent Heuveline |
| 264 | 1 | |c 2017 | |
| 300 | |a 12 | ||
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Gesehen am 01.10.2018 | ||
| 520 | |a We investigate time and energy to solution for the CPU- and GPU-based execution of the compute intensive smoother and grid transfer operators in a geometric multigrid linear solver. We use a hybrid parallel implementation for both shared and distributed memory multi-core host systems comprising CUDA-capable devices. Our numerical experiments are designed to assess the effect of combining an MPI-parallel multigrid framework with OpenMP host threads or CUDA accelerators instead of MPI-only CPU computations for various parallel setups. We present runtime and energy measurements from a quad-CPU test system equipped with two GPUs. We find that using an accelerated asynchronous smoother can yield substantial savings of time and energy to solution over using a host-only Jacobi smoother in small and medium sized host systems with one or two multi-core CPUs. The acceleration of the grid transfer operators also yields a benefit, yet smaller than the benefit from the smoother. For large host systems a hybrid MPI-OpenMP parallelization turns out to be most beneficial with respect to energy consumption, although it is not the fastest option. | ||
| 534 | |c 2016 | ||
| 650 | 4 | |a Energy-aware numerics | |
| 650 | 4 | |a Geometric multigrid | |
| 650 | 4 | |a Heterogeneous platforms | |
| 650 | 4 | |a Hybrid parallelization | |
| 650 | 4 | |a Performance and energy assessment | |
| 700 | 1 | |a Heuveline, Vincent |d 1968- |e VerfasserIn |0 (DE-588)1046579266 |0 (DE-627)776691880 |0 (DE-576)399904727 |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Journal of parallel and distributed computing |d Amsterdam [u.a.] : Elsevier, 1984 |g 100(2017), Seite 181-192 |h Online-Ressource |w (DE-627)267328222 |w (DE-600)1469781-6 |w (DE-576)104193921 |7 nnas |a Energy-efficient multigrid smoothers and grid transfer operators on multi-core and GPU clusters |
| 773 | 1 | 8 | |g volume:100 |g year:2017 |g pages:181-192 |g extent:12 |a Energy-efficient multigrid smoothers and grid transfer operators on multi-core and GPU clusters |
| 856 | 4 | 0 | |u http://dx.doi.org/10.1016/j.jpdc.2016.05.006 |x Verlag |x Resolving-System |3 Volltext |
| 856 | 4 | 0 | |u http://www.sciencedirect.com/science/article/pii/S0743731516300363 |x Verlag |3 Volltext |
| 951 | |a AR | ||
| 992 | |a 20181001 | ||
| 993 | |a Article | ||
| 994 | |a 2017 | ||
| 998 | |g 1046579266 |a Heuveline, Vincent |m 1046579266:Heuveline, Vincent |d 700000 |d 704000 |e 700000PH1046579266 |e 704000PH1046579266 |k 0/700000/ |k 1/700000/704000/ |p 2 |y j | ||
| 998 | |g 1031120645 |a Wlotzka, Martin |m 1031120645:Wlotzka, Martin |d 700000 |d 704000 |e 700000PW1031120645 |e 704000PW1031120645 |k 0/700000/ |k 1/700000/704000/ |p 1 |x j | ||
| 999 | |a KXP-PPN1581442211 |e 3027469983 | ||
| BIB | |a Y | ||
| SER | |a journal | ||
| JSO | |a {"origin":[{"dateIssuedKey":"2017","dateIssuedDisp":"2017"}],"relHost":[{"physDesc":[{"extent":"Online-Ressource"}],"note":["Gesehen am 14.08.2020"],"origin":[{"publisherPlace":"Amsterdam [u.a.] ; New York, NY [u.a.]","dateIssuedDisp":"1984-","dateIssuedKey":"1984","publisher":"Elsevier ; Academic Press"}],"id":{"zdb":["1469781-6"],"eki":["267328222"]},"recId":"267328222","disp":"Energy-efficient multigrid smoothers and grid transfer operators on multi-core and GPU clustersJournal of parallel and distributed computing","title":[{"title":"Journal of parallel and distributed computing","title_sort":"Journal of parallel and distributed computing"}],"part":{"year":"2017","extent":"12","pages":"181-192","text":"100(2017), Seite 181-192","volume":"100"},"pubHistory":["1.1984 - 74.2014; Vol. 75.2015 -"],"type":{"media":"Online-Ressource","bibl":"periodical"},"language":["eng"]}],"id":{"doi":["10.1016/j.jpdc.2016.05.006"],"eki":["1581442211"]},"physDesc":[{"extent":"12 S."}],"person":[{"roleDisplay":"VerfasserIn","family":"Wlotzka","given":"Martin","role":"aut","display":"Wlotzka, Martin"},{"roleDisplay":"VerfasserIn","family":"Heuveline","given":"Vincent","display":"Heuveline, Vincent","role":"aut"}],"note":["Gesehen am 01.10.2018"],"name":{"displayForm":["Martin Wlotzka, Vincent Heuveline"]},"type":{"bibl":"article-journal","media":"Online-Ressource"},"language":["eng"],"recId":"1581442211","title":[{"title_sort":"Energy-efficient multigrid smoothers and grid transfer operators on multi-core and GPU clusters","title":"Energy-efficient multigrid smoothers and grid transfer operators on multi-core and GPU clusters"}]} | ||
| SRT | |a WLOTZKAMARENERGYEFFI2017 | ||