AUTHOREA
Log in Sign Up Browse Preprints
LOG IN SIGN UP
Yuyao Kong
Yuyao Kong

Public Documents 1
A time-domain compute-in-memory architecture based on pulse width modulation for mult...
Yuyao Kong
Xin Si

Yuyao Kong

and 1 more

January 13, 2023
Compute-in-memory (CIM) architecture is an effective way to reduce the energy efficiency of convolutional neural networks (DNNs). This letter presents a time-domain compute-in-memory design based on pulse width modulation. Unique-weight convolution method is proposed for multi-bits convolutional operations. A time-charge domain quantizer is also presented to quantify the computation pulses of multi-rows in parallel. Fabricated in 28nm CMOS technology, this design achieves 56.9% and 67.3% accuracy, 59.84Tops/W and 67.45Tops/W energy efficiency for 1-8-b AlexNet and VGG16, respectively.

| Powered by Authorea.com

  • Home