Abstract | ||
---|---|---|
While Speculative Multithreading (SM) on a Chip Mul- tiprocessor (CMP) has the ability to speed-up hard-to- parallelize applications, the power inefficiency of aggres- sive speculation is a concern. To improve SM's power effi- ciency, we note that not all the tasks that are running in a SM environment are equally critical. To leverage this insight, this paper develops a novel, widely-applicable task-criticality model for SM. It also pro- posesCAP, a novel architecture that builds a task-criticality graph dynamically and uses it to make scheduling decisions in a SM CMP. Experiments with SPECint, SPECfp, and Olden applications show that, in a CMP with one fast core and three slow ones, the E ◊ D2 with CAP is, on average, 91-95% of that without. Moreover, it is only 77-91% of the E◊D2 of a CMP with four fast cores and no CAP. Overall, we argue that scheduling for task criticality is beneficial. |
Year | DOI | Venue |
---|---|---|
2007 | 10.1109/ICCD.2007.4601932 | ICCD |
Keywords | Field | DocType |
multi-threading,multiprocessing systems,scheduling,CAP,chip multiprocessor,criticality analysis,power-efficient speculative multithreading,scheduling,task-criticality graph | Multithreading,Scheduling (computing),Computer science,SPECfp,Parallel computing,Speculative multithreading,Real-time computing,Multiprocessing,SPECint,Dynamic priority scheduling,Out-of-order execution,Embedded system | Conference |
ISSN | Citations | PageRank |
1063-6404 | 6 | 0.43 |
References | Authors | |
12 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
James Tuck | 1 | 564 | 33.06 |
Wei Liu | 2 | 1124 | 52.28 |
Josep Torrellas | 3 | 3838 | 262.89 |