An 82nW 0.53pJ/SOP Clock-Free Spiking Neural Network with 40µs Latency for AloT Wake-Up Functions Using Ultimate-Event-Driven Bionic Architecture and Computing-in-Memory Technique.
Ying LiuZhixuan WangWei HeLinxiao ShenYihan ZhangPeiyu ChenMeng WuHao ZhangPeng ZhouJinguang LiuGuangyu SunJiayoon RuLe YeRu HuangPublished in: ISSCC (2022)
Keyphrases
- event driven
- cerebellar model
- spiking neural networks
- publish subscribe
- biologically inspired
- information delivery
- real time
- artificial neural networks
- feed forward
- associative memory
- liquid state machine
- spiking neurons
- memory access
- biological systems
- basis functions
- control system
- data model
- multiscale
- metadata
- genetic algorithm
- machine learning