RT Conference Proceedings
T1 Resource Management Optimization in Multi-Processor Platforms
A1 Hristov, Atanas
A1 Nikolova, Iva
A1 Zapryanov, Georgi
A1 Kimovski, Dragi
A1 Kumbaroska, Vesna
A2 Carretero Pérez, Jesús
A2 García Blas, Javier
A2 Margenov, Svetozar
A2 Universidad Carlos III de Madrid. Computer Architecture, Communications and Systems Group (ARCOS)
AB The modern high-performance computing systems (HPCS) are composed of hundreds of thousand computational nodes. Aneffective resource allocation in HPCS is a subject for many scientific research investigations. Many programming models foreffective resources allocation have been proposed. The main purpose of those models is to increase the parallel performanceof the HPCS. This paper investigates the efficiency of parallel algorithm for resource management optimization based onArtificial Bee Colony (ABC) metaheuristic while solving a package of NP-complete problems on multi-processor platform.Inorder to achieve minimal parallelization overhead in each cluster node, a multi-level hybrid programming model is proposed thatcombines coarse-grain and fine-grain parallelism. Coarse-grain parallelism is achieved through domain decomposition by messagepassing among computational nodes using Message Passing Interface (MPI) and fine-grain parallelism is obtained by loop-levelparallelism inside each computation node by compiler-based thread parallelization via Intel TBB. Parallel communicationsprofiling is made and parallel performance parameters are evaluated on the basis of experimental results.
SN 978-84-617-7450-0
YR 2016
FD 2016-12
LK https://hdl.handle.net/10016/24230
UL https://hdl.handle.net/10016/24230
LA eng
NO Proceedings of: Third International Workshop on Sustainable Ultrascale Computing Systems (NESUS 2016). Sofia (Bulgaria), October, 6-7, 2016.
DS e-Archivo
RD 30 jun. 2024