ARETLGNEDec 10, 2020

A Custom 7nm CMOS Standard Cell Library for Implementing TNN-based Neuromorphic Processors

arXiv:2012.05419v2
AI Analysis

This work provides a highly energy-efficient hardware implementation solution for neuromorphic processors based on TNNs, beneficial for edge AI applications requiring low power and small form factor.

This paper presents a custom 7nm CMOS standard cell library with optimized macro extensions for implementing Temporal Neural Networks (TNNs). A TNN prototype for MNIST, comprising 13,750 neurons and 315,000 synapses, achieved a compact die area of 1.56mm2 and consumed only 1.69mW.

A set of highly-optimized custom macro extensions is developed for a 7nm CMOS cell library for implementing Temporal Neural Networks (TNNs) that can mimic brain-like sensory processing with extreme energy efficiency. A TNN prototype (13,750 neurons and 315,000 synapses) for MNIST requires only 1.56mm2 die area and consumes only 1.69mW.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes