Search papers, labs, and topics across Lattice.
NeuDW-CIM is a neuromorphic Compute-in-Memory macro designed for Spiking Neural Networks, fabricated in a 65 nm CMOS process, achieving significant energy efficiency of 0.8 pJ/SOP. The macro features a custom twin 9T bit-cell for handling ternary inputs and weights, and employs a reconfigurable non-linear In-Memory ADC to emulate biological dendritic functions, resulting in high accuracies of 97.2% on N-MNIST and 95.5% on DVS Gesture. Additionally, its Top-K Winner mode enhances performance by reducing latency and improving energy efficiency through sparse updates, marking a 1.6x improvement over previous designs.
Achieving 97.2% accuracy on N-MNIST while consuming only 0.8 pJ/SOP could redefine energy-efficient neuromorphic computing.
This work presents NeuDW-CIM, a highly efficient neuromorphic Compute-in-Memory (CIM) macro for Spiking Neural Networks (SNNs) implemented in 65 nm CMOS. The design introduces a custom twin 9T bit-cell for ternary in-puts/weights and a reconfigurable non-linear In-Memory ADC (IMA). The macro supports two specialized modes: 1) Nonlinear Dendrite (NLD) mode, which utilizes reconfigurable IMA to emulate biological dendritic functions, achieving measured accuracies of 97.2% on N-MNIST and 95.5% on DVS Gesture; and 2) Top-K Winner (KWN) mode, featuring an early-stopping mechanism that reduces IMA conversion latency by 30% and digital LIF latency by 10x. Benefiting from the sparse update in KWN mode, NeuDW-CIM achieves a measured energy efficiency (EE) of 0.8 pJ/SOP (1.6x improvement).
