Scheduling of Linear Algebra Kernels on Multiple Heterogeneous Resources
hal.structure.identifier | Reformulations based algorithms for Combinatorial Optimization [Realopt] | |
dc.contributor.author | BEAUMONT, Olivier | |
hal.structure.identifier | STatic Optimizations, Runtime Methods [STORM] | |
dc.contributor.author | COJEAN, Terry | |
hal.structure.identifier | Reformulations based algorithms for Combinatorial Optimization [Realopt] | |
dc.contributor.author | EYRAUD-DUBOIS, Lionel | |
hal.structure.identifier | High-End Parallel Algorithms for Challenging Numerical Simulations [HiePACS] | |
dc.contributor.author | GUERMOUCHE, Abdou | |
hal.structure.identifier | STatic Optimizations, Runtime Methods [STORM] | |
dc.contributor.author | KUMAR, Suraj | |
dc.date.accessioned | 2024-04-04T03:13:50Z | |
dc.date.available | 2024-04-04T03:13:50Z | |
dc.date.issued | 2016-12 | |
dc.date.conference | 2016-12-19 | |
dc.identifier.uri | https://oskar-bordeaux.fr/handle/20.500.12278/193986 | |
dc.description.abstractEn | In this paper, we consider task-based dense linear algebra applications on a single heterogeneous node which contains regular CPU cores and a set of GPU devices. Efficient scheduling strategies are crucial in this context in order to achieve good and portable performance. HeteroPrio, a resource-centric dynamic scheduling strategy has been introduced in a previous work and evaluated for the special case of nodes with exactly two types of resources. However, this restriction can be limiting, for example on nodes with several types of accelerators, but not only this. Indeed, an interesting approach to increase resource usage is to group several CPU cores together, which allows to use intra-task parallelism. We propose a generalization of HeteroPrio to the case with several classes of heterogeneous workers. We provide extensive evaluation of this algorithm with Cholesky factorization, both through simulation and actual execution, compared with HEFT-based scheduling strategy, the state of the art dynamic scheduling strategy for heterogeneous systems. Experimental evaluation shows that our approach is efficient even for highly heterogeneous configurations and significantly outperforms HEFT-based strategy. | |
dc.description.sponsorship | Solveurs pour architectures hétérogènes utilisant des supports d'exécution - ANR-13-MONU-0007 | |
dc.language.iso | en | |
dc.publisher | IEEE | |
dc.subject.en | Cholesky Factorization | |
dc.subject.en | StarPU | |
dc.subject.en | Resource Aggregation | |
dc.subject.en | Simulation | |
dc.subject.en | Task-based Scheduling | |
dc.subject.en | Heterogeneous Platforms | |
dc.subject.en | Linear Algebra | |
dc.title.en | Scheduling of Linear Algebra Kernels on Multiple Heterogeneous Resources | |
dc.type | Communication dans un congrès | |
dc.identifier.doi | 10.1109/HiPC.2016.045 | |
dc.subject.hal | Informatique [cs] | |
bordeaux.hal.laboratories | Institut de Mathématiques de Bordeaux (IMB) - UMR 5251 | * |
bordeaux.institution | Université de Bordeaux | |
bordeaux.institution | Bordeaux INP | |
bordeaux.institution | CNRS | |
bordeaux.conference.title | International Conference on High Performance Computing, Data, and Analytics (HiPC 2016) | |
bordeaux.country | IN | |
bordeaux.conference.city | Hyderabad | |
bordeaux.peerReviewed | oui | |
hal.identifier | hal-01361992 | |
hal.version | 1 | |
hal.invited | non | |
hal.proceedings | oui | |
hal.conference.end | 2016-12-22 | |
hal.popular | non | |
hal.audience | Internationale | |
hal.origin.link | https://hal.archives-ouvertes.fr//hal-01361992v1 | |
bordeaux.COinS | ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.date=2016-12&rft.au=BEAUMONT,%20Olivier&COJEAN,%20Terry&EYRAUD-DUBOIS,%20Lionel&GUERMOUCHE,%20Abdou&KUMAR,%20Suraj&rft.genre=unknown |
Files in this item
Files | Size | Format | View |
---|---|---|---|
There are no files associated with this item. |