Abstract | ||
---|---|---|
In this paper we present a framework for automatic detection and application of the best binding between threads of a running parallel application and processor cores in a shared memory system, by making use of hardware performance counters. This is especially important within the scope of multicore architectures with shared cache levels. We demonstrate that many applications from the SPEC OMP benchmark show quite sensitive runtime behavior depending on the thread/core binding used. In our tests, the proposed framework is able to find the best binding in nearly all cases. The proposed framework is intended to supplement job scheduling systems for better automatic exploitation of systems with multicore processors, as well as making programmers aware of this issue by providing measurement logs. |
Year | DOI | Venue |
---|---|---|
2011 | 10.1007/978-3-642-19448-1_12 | T. HiPEAC |
Keywords | Field | DocType |
automatic detection,parallel application,multicore processor,multicore architecture,spec omp benchmark,automated optimization,thread-to-core pinning,shared memory system,proposed framework,shared cache level,better automatic exploitation,hardware performance counter,multicore system,job scheduling,multicore processors | Shared memory,Computer science,Parallel computing,Real-time computing,Thread (computing),Job scheduler,Spec#,Multi-core processor,Embedded system,Multicore systems | Journal |
Volume | Citations | PageRank |
3 | 38 | 1.61 |
References | Authors | |
10 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Tobias Klug | 1 | 93 | 9.29 |
Michael Ott | 2 | 43 | 2.19 |
Josef Weidendorfer | 3 | 115 | 17.98 |
Carsten Trinitis | 4 | 151 | 29.80 |