All-Digital Time-Domain Compute-in-Memory Engine for Binary Neural Networks With 1.05 POPS/W Energy Efficiency

This paper presents an all-digital time-domain compute-in-memory (TDCIM) engine for binary neural networks (BNNs), which is based on commercial standard cells facilitating technology mapping. The proposed TDCIM engine exploits energy-efficient computing principles, supports data reuse and employs do...

Full description

Saved in:
Bibliographic Details
Published inESSCIRC 2022- IEEE 48th European Solid State Circuits Conference (ESSCIRC) pp. 149 - 152
Main Authors Lou, Jie, Lanius, Christian, Freye, Florian, Stadtmann, Tim, Gemmeke, Tobias
Format Conference Proceeding
LanguageEnglish
Published IEEE 19.09.2022
Subjects
Arrays
binary neural network
commercial standard cells
compute-in-memory
double-edge operation
Neural networks
Routing
Solid state circuits
Systematics
Throughput
Time measurement
time-domain
wave-pipelining
Online AccessGet full text

Cover

More Information
Summary:This paper presents an all-digital time-domain compute-in-memory (TDCIM) engine for binary neural networks (BNNs), which is based on commercial standard cells facilitating technology mapping. The proposed TDCIM engine exploits energy-efficient computing principles, supports data reuse and employs double-edge triggered operation. Time domain wave-pipelining technique is also introduced to improve throughput by 1.5x while preserving accuracy. We use Structured Data-Path (SDP) placement and custom routing flow during place and route (P&R) to reduce systematic variations. The measured arrival time of different MAC results is sufficiently bounded to preserve accuracy across PVT variations. Fabricated in a 22nm process, the proposed BNN engine can achieve an energy efficiency of 1.05 POPS/W at 0.5V matching the accuracy of the PyTorch baseline of 99.14% on the MNIST dataset.
DOI:10.1109/ESSCIRC55480.2022.9911382