16.7 A 40-310TOPS/W SRAM-Based All-Digital Up to 4b In-Memory Computing Multi-Tiled NN Accelerator in FD-SOI 18nm for Deep-Learning Edge Applications

Winter 2022–Spring 2023: Looking Back at Switzerland Chapter Activity [Chapters]

“…Switzerland C h a pter members coauthored seven submissions, representing a good mix of industry, corporate research, and academia; Sony Advanced Visual Sensing, Zurich, Switzerland, contrib-uted two papers [3], [4]. For ISSCC session 16, "Efficient Compute-In-Memory-Based Processors for ML," Thomas Boesch, ST Microelectronics, contributed to [5], showing a nice neural network accelerator leveraging in-memory computing on 18-nm FDSOI nodes. For session 17, "High-Speed Data Converters," researchers from ETH Zurich and IBM's Zurich research laboratory contributed [6], with technical achievements in 5-nm technology.…”