Fully-Asynchronous Cache-Efficient Simulation of Detailed Neural Networks
Magalhães B, Sterling T, Hines M, Schürmann F. Fully-Asynchronous Cache-Efficient Simulation of Detailed Neural Networks. Lecture Notes In Computer Science 2019, 11538: 421-434. DOI: 10.1007/978-3-030-22744-9_33.Peer-Reviewed Original ResearchRuntime systemNeural networkScientific applicationsLarge-scale scientific applicationsAsynchronous runtime systemHPX runtime systemParalleX execution modelOverlap of computationLinear data structuresBetter cache localityCompute architecturesExecution modelParallel executionCore kernelsData structureCache localitySynchronous executionMemory spaceCache levelsNode levelBenchmark resultsExecutionLower timeNumber of timestepsNetwork