IMAGINE: An 8-to-1b 22nm FD-SOI Compute-In-Memory CNN Accelerator With an End-to-End Analog Charge-Based 0.15-8POPS/W Macro Featuring Distribution-Aware Data Reshaping

Adrian Kneip, Martin Lefebvre, Pol Maistriaux +1

Charge-domain compute-in-memory (CIM) SRAMs have recently become an enticing compromise between computing efficiency and accuracy to process sub-8b convolutional neural networks (C…