Skip to Main content Skip to Navigation
Conference papers

Data-Aware Task Scheduling on Multi-Accelerator based Platforms

Cédric Augonnet 1, 2 Jérôme Clet-Ortega 1, 2 Samuel Thibault 1, 2 Raymond Namyst 1, 2 
2 RUNTIME - Efficient runtime systems for parallel architectures
Inria Bordeaux - Sud-Ouest, UB - Université de Bordeaux, CNRS - Centre National de la Recherche Scientifique : UMR5800
Abstract : To fully tap into the potential of heterogeneous machines composed of multicore processors and multiple accelerators, simple offloading approaches in which the main trunk of the application runs on regular cores while only specific parts are offloaded on accelerators are not sufficient. The real challenge is to build systems where the application would permanently spread across the entire machine, that is, where parallel tasks would be dynamically scheduled over the full set of available processing units. To face this challenge, we previously proposed StarPU, a runtime system capable of scheduling tasks over multicore machines equipped with GPU accelerators. StarPU uses a software virtual shared memory (VSM) that provides a high-level programming interface and automates data transfers between processing units so as to enable a dynamic scheduling of tasks. We now present how we have extended StarPU to minimize the cost of transfers between processing units in order to efficiently cope with multi-GPU hardware configurations. To this end, our runtime system implements data prefetching based on asynchronous data transfers, and uses data transfer cost prediction to influence the decisions taken by the task scheduler. We demonstrate the relevance of our approach by benchmarking two parallel numerical algorithms using our runtime system. We obtain significant speedups and high efficiency over multicore machines equipped with multiple accelerators. We also evaluate the behaviour of these applications over clusters featuring multiple GPUs per node, showing how our runtime system can combine with MPI.
Document type :
Conference papers
Complete list of metadata

Cited literature [15 references]  Display  Hide  Download
Contributor : Cédric Augonnet Connect in order to contact the contributor
Submitted on : Wednesday, October 6, 2010 - 4:00:29 PM
Last modification on : Saturday, June 25, 2022 - 10:32:18 AM
Long-term archiving on: : Thursday, October 25, 2012 - 4:35:23 PM


Files produced by the author(s)


  • HAL Id : inria-00523937, version 1



Cédric Augonnet, Jérôme Clet-Ortega, Samuel Thibault, Raymond Namyst. Data-Aware Task Scheduling on Multi-Accelerator based Platforms. 16th International Conference on Parallel and Distributed Systems, Dec 2010, Shangai, China. ⟨inria-00523937⟩



Record views


Files downloads