Mostra i principali dati dell'item
Runtime scheduling of dynamic parallelism on accelerator-based multi-core systems
dc.creator | Blagojevic, F. | en |
dc.creator | Nikolopoulos, D. S. | en |
dc.creator | Stamatakis, A. | en |
dc.creator | Antonopoulos, C. D. | en |
dc.creator | Curtis-Maury, M. | en |
dc.date.accessioned | 2015-11-23T10:23:54Z | |
dc.date.available | 2015-11-23T10:23:54Z | |
dc.date.issued | 2007 | |
dc.identifier | 10.1016/j.parco.2007.09.004 | |
dc.identifier.issn | 0167-8191 | |
dc.identifier.uri | http://hdl.handle.net/11615/26314 | |
dc.description.abstract | We explore runtime mechanisms and policies for scheduling dynamic multi-grain parallelism on heterogeneous multicore processors. Heterogeneous multi-core processors integrate conventional cores that run legacy codes with specialized cores that serve as computational accelerators. The term multi-grain parallelism refers to the exposure of multiple dimensions of parallelism from within the runtime system, so as to best exploit a parallel architecture with heterogeneous computational capabilities between its cores and execution units. We investigate user-level schedulers that dynamically "rightsize" the dimensions and degrees of parallelism on the cell broadband engine. The schedulers address the problem of mapping application-specific concurrency to an architecture with multiple hardware layers of parallelism, without requiring programmer intervention or sophisticated compiler support. We evaluate recently introduced schedulers for event-driven execution and utilization-driven dynamic multi-grain parallelization on Cell. We also present a new scheduling scheme for dynamic multi-grain parallelism, S-MGPS, which uses sampling of dominant execution phases to converge to the optimal scheduling algorithm. We evaluate S-MGPS on an IBM Cell BladeCenter with two realistic bioinformatics applications that infer large phylogenies. S-MGPS performs within 2-10% of the optimal scheduling algorithm in these applications, while exhibiting low overhead and little sensitivity to application-dependent parameters. (c) 2007 Elsevier B.V. All rights reserved. | en |
dc.source | Parallel Computing | en |
dc.source.uri | <Go to ISI>://WOS:000251752900005 | |
dc.subject | heterogeneous multi-core processors | en |
dc.subject | accelerator-based parallel | en |
dc.subject | architectures | en |
dc.subject | runtime systems for parallel programming | en |
dc.subject | Cell broadband | en |
dc.subject | engine | en |
dc.subject | Computer Science, Theory & Methods | en |
dc.title | Runtime scheduling of dynamic parallelism on accelerator-based multi-core systems | en |
dc.type | journalArticle | en |
Files in questo item
Files | Dimensione | Formato | Mostra |
---|---|---|---|
Nessun files in questo item. |