[23a-M206-1~10] FS.1 Focused Session "AI Electronics"

△ [23a-M206-7] Non-Volatile Memory & SRAM Hybrid Computation-in-Memory for Reservoir Computing Processing Event Data

〇Shunsuke Koshino¹, Chihiro Matsui¹, Ken Takeuchi¹ (1.Univ. Tokyo)

Keywords:Reservoir Computing

Non-Volatile Memory (NVM) & SRAM Hybrid Computation-in-Memory (CiM) for Reservoir Computing is proposed. This work is the proposal to recognize event data only by Reservoir Computing. Proposed Reservoir Computing achieves 87.9% accuracy for Event-based Vision Sensor (EVS) gesture dataset with prompt learning. With a detailed memory error analysis, proposed Reservoir Computing can tolerate 0.1% bit-error rate (BER) for both EVS gesture and MNIST. NVM & SRAM Hybrid CiM is proposed, which optimizes the memory type based on the error tolerance of each layer’s weights. The memory cell array area of CiM decreases by 89.2% and 0.1% BER of NVM is acceptable.