Tempo-CIM: A RRAM Compute-in-Memory Neuromorphic Accelerator With Area-Efficient LIF Neuron and Split-Train-Merged-Inference Algorithm for Edge AI Applications.
Jingwen JiangKeji ZhouJinhao LiangFengshi TianChenyang ZhaoJianguo YangXiaoyong XueXiaoyang ZengPublished in: IEEE J. Emerg. Sel. Topics Circuits Syst. (2023)
Keyphrases
- single pass
- optimization algorithm
- memory usage
- high efficiency
- memory requirements
- computationally efficient
- learning algorithm
- memory space
- parallel implementation
- detection algorithm
- search space
- objective function
- undirected graph
- optimal solution
- computational complexity
- highly efficient
- dynamic programming
- cost function
- np hard
- pruning strategy
- edge detection algorithm
- matching algorithm
- particle swarm optimization
- inference mechanism
- clustering algorithm
- neural network
- weighted graph
- lower bound
- simulated annealing
- worst case
- probabilistic model
- computational cost