IBM NorthPole: An Architecture for Neural Network Inference with a 12nm ChipAndrew S. CassidyJohn V. Arthuret al.2024ISSCC 2024
07 Mar 2022US11270196Multi-mode Low-precision Inner-product Computation Circuits For Massively Parallel Neural Inference Engine
10 Aug 2021GB2586556Time, Space, And Energy Efficient Neural Inference Via Parallelism And On-chip Memory