Di Yang
ABSTRACT
In order to give full play to the multi-core parallel computing ability of CPU+GPU heterogeneous processing platform and maximize the efficient utilization of platform resources, this paper analyses the architecture characteristics of the platform, improves the HEFT algorithm, and proposes a load deployment decision algorithm. Firstly, the algorithm pre judges the load, establishes the scheduling queue from large to small, and pre allocates the task to GPU and CPU; then it uses HEFT algorithm to sort the tasks, but in the processor selection stage, the algorithm only selects the corresponding GPU and CPU from the two scheduling queues. Simulation results show that, compared with the classic HEFT algorithm, the load deployment decision algorithm greatly improves the utilization rate of GPU and the speedup ratio in the calculation process. It is a more load balanced algorithm with shorter scheduling length and higher efficiency.
- ATEF A, HAGRAS T, MAHDY Y B, Lower-bound complexity algorithm for task scheduling on heterogeneous grid[J]. Computing, 2017, 99(11):1125-1145.Google Scholar
Digital Library
- ZHANG W, HU Y, HE H, Linear and dynamic programming algorithms for real-time task scheduling with task duplication[J]. Journal of supercomputing, 2017(1):1-16.Google Scholar
- SHIN I, EASWARAN A, LEE I. Hierarchical scheduling framework for virtual clustering of multiprocessors[C]// Euromicro conference on real-time systems. IEEE Computer Society, 2008:181-190.Google Scholar
- TSAI Y L, LIU H C, HUANG K C. Adaptive dual-criteria task group allocation for clustering-based multi-workflow scheduling on parallel computing platform[J]. Journal of supercomputing, 2015, 71(10):3811-3831.Google ScholarDigital Library
- MU. Advanced list scheduling heuristic for task scheduling with communication contention for parallel embedded systems[J]. Science China (Information sciences), 2010, 53(11):2272-2286.Google Scholar
- MU P, NEZAN J F, RAULET M. A list scheduling heuristic with new node priorities and critical child technique for task scheduling with communication contention [J]. Lecture notes in electrical engineering, 2013, 73:217-236.Google Scholar
- ALEBRAHIM S, AHMAD I. Task scheduling for heterogeneous computing systems [J]. Journal of supercomputing, 2017, 73(6):2313-2338.Google ScholarDigital Library
- TOPCUOGLU H, HARIRI S, WU M Y. Performance-effective and low-complexity task scheduling for heterogeneous computing [J]. IEEE transactions on parallel & distributed systems, 2002, 13(3):260-274.Google ScholarDigital Library
- CANON L-C, JEANNOT E, SAKELLARIOU R, Comparative evaluation of the robustness of dag scheduling heuristics[A]. Proceedings of grid computing, Achievements and Prospects[C]. Springer, 2008: 73-84.Google Scholar
- Yang Di, Ma Jinquan, Yue Chunsheng, Longest path list scheduling algorithm for heterogeneous processing platforms [J]. Journal of University of information engineering, 2021 (S2): 3Google Scholar
- Wang Xuecheng. Research on task scheduling technology based on CPU+GPU heterogeneous signal processing platform [D]. Zhengzhou: University of information engineering, 2019Google Scholar
Recommendations
Adaptive Optimization for Petascale Heterogeneous CPU/GPU Computing
CLUSTER '10: Proceedings of the 2010 IEEE International Conference on Cluster ComputingIn this paper, we describe our experiment developing an implementation of the Linpack benchmark for TianHe-1, a petascale CPU/GPU supercomputer system, the largest GPU-accelerated system ever attempted before. An adaptive optimization framework is ...
AOA: Adaptive Overclocking Algorithm on CPU-GPU Heterogeneous Platforms
Algorithms and Architectures for Parallel ProcessingAbstractAlthough GPUs have been used to accelerate various convolutional neural network algorithms with good performance, the demand for performance improvement is still continuously increasing. CPU/GPU overclocking technology brings opportunities for ...
Resource-efficient utilization of CPU/GPU-based heterogeneous supercomputers for Bayesian phylogenetic inference
Bayesian inference is one of the most important methods for estimating phylogenetic trees in bioinformatics. Due to the potentially huge computational requirements, several parallel algorithms of Bayesian inference have been implemented to run on CPU-...
Comments