Tiny but mighty
Marcelo Orenes-Vera, Aninda Manocha, Jonathan Balkind, Fei Gao, Juan L. Aragón, David Wentzlaff, Margaret Martonosi
Abstract
Modern computing systems employ significant heterogeneity and specialization to meet performance targets at manageable power. However, memory latency bottlenecks remain problematic, particularly for sparse neural network and graph analytic applications where indirect memory accesses (IMAs) challenge the memory hierarchy.
Topics & Concepts
Computer scienceMemory hierarchyTheoretical computer scienceGraphCAS latencyLatency (audio)Parallel computingHierarchyArtificial neural networkDistributed computingComputer architectureArtificial intelligenceSemiconductor memoryComputer hardwareMemory controllerCacheTelecommunicationsEconomicsMarket economyParallel Computing and Optimization TechniquesCloud Computing and Resource ManagementAdvanced Data Storage Technologies