2013 · 13 citations · 37 references
ComputingComputer EngineeringComputer ScienceAlgorithmsMachine LearningParallel ComputingMulticore Processor DesignSpeculative ExecutionAutomatic ParallelizationHeuristic ApproachMultithreadingParallelizationPartitioning StrategiesProcessor PartitioningSpeculative MultithreadingWorkload Partitioning
Speculative multithreading (SpMT) is a thread-level automatic parallelization technique to accelerate sequential programs on multi-core. The existing heuristic-based approaches are only suitable for one kind of programs and cannot guarantee to get the optimal solution of thread partitioning. In this paper, we propose a novel thread partitioning approach based on machine learning to partition irregular programs into multithreads. It mainly includes: generating sufficient training samples, building and applying the prediction model to partition the irregular programs. By using the thread partition approach, an unseen irregular program can obtain a stable, much higher speedup than the heuristic-based approaches. On the Prophet, which is a SpMT processor to evaluate the performance of multithreaded programs, the novel thread partitioning approach is evaluated and reaches an average speedup of 1.80 on 4-core processor. Experiments show that our proposed approach can obtain a significant increase in speedup and Olden benchmarks deliver a better performance improvement of 5.41% than the traditional heuristic-based approach.
37
Pattern Recognition and Machine Learning
Kybernetes · 2007 · 8.4K citations
Gurindar S. Sohi, Scott E. Breach, T. N. Vijaykumar · 1995 · 978 citations
Software-based replication for fault tolerance
Rachid Guerraoui, A. Schiper · Computer · 1997 · 372 citations · Full text
Lance Hammond, Benedict A. Hubbert, M. Stewart Siu et al. · IEEE Micro · 2000 · 342 citations
Wei Liu, James Tuck, Luís Ceze et al. · 2006 · 216 citations