Proceedings of the 59th ACM/IEEE Design Automation Conference · 2022 · 18 citations · 8 references
ComputingComputer ScienceInformation RetrievalMachine LearningNatural Language ProcessingDeep LearningInformation ExtractionRepresentation LearningSemantic SearchText MiningTopic ModelingData DiscoveryDeep Feature LearningLarge-scale Natural Language ProcessingSchema MatchingSemantic Relation ExtractionSemantic Relations ExtractionSemantic TypesTable Information Extraction
Event-driven spiking neural networks (SNNs) have shown great promise for being strikingly energy-efficient. SNN neurons integrate the spikes, accumulate the membrane potential, and fire output spike when the potential exceeds a threshold. Existing SNN accelerators, however, have to carry out such accumulation-comparison operation in serial. Repetitive spike generation at each time step not only increases latency as well as overall energy budget, but also incurs memory access overhead of fetching membrane potentials, both of which lessen the efficiency of SNN accelerators. Meanwhile, inherent highly sparse spikes of SNNs lead to imbalanced workloads among neurons that hurdle the utilization of processing elements (PEs).
8
Yu‐Hsin Chen, Joel Emer, Vivienne Sze · ACM SIGARCH Computer Architecture News · 2016 · 1.2K citations
Rethinking the performance comparison between SNNS and ANNS
Lei Deng, Yujie Wu, Xing Hu et al. · Neural Networks · 2019 · 333 citations
SpinalFlow: An Architecture and Dataflow Tailored for Spiking Neural Networks
Surya Narayanan, Karl Taht, Rajeev Balasubramonian et al. · 2020 · 113 citations