CAP-RAM: A Charge-Domain In-Memory Computing 6T-SRAM for Accurate and Precision-Programmable CNN Inference

TL;DR

CAP-RAM introduces a charge-domain in-memory computing SRAM macro for energy-efficient, high-accuracy CNN inference with programmable bitwidths, leveraging novel circuitry and architecture for superior linearity and throughput.

Contribution

It presents a novel charge-domain MAC mechanism and a semi-parallel architecture enabling accurate, programmable CNN inference in SRAM with improved energy efficiency.

Findings

Achieves 98.8% accuracy on MNIST with 512x128 macro.

Supports up to six bit-width levels for weights.

Attains 49.4 TOPS/W energy efficiency.

Abstract

A compact, accurate, and bitwidth-programmable in-memory computing (IMC) static random-access memory (SRAM) macro, named CAP-RAM, is presented for energy-efficient convolutional neural network (CNN) inference. It leverages a novel charge-domain multiply-and-accumulate (MAC) mechanism and circuitry to achieve superior linearity under process variations compared to conventional IMC designs. The adopted semi-parallel architecture efficiently stores filters from multiple CNN layers by sharing eight standard 6T SRAM cells with one charge-domain MAC circuit. Moreover, up to six levels of bit-width of weights with two encoding schemes and eight levels of input activations are supported. A 7-bit charge-injection SAR (ciSAR) analog-to-digital converter (ADC) getting rid of sample and hold (S&H) and input/reference buffers further improves the overall energy efficiency and throughput. A 65-nm prototype validates the excellent linearity and computing accuracy of CAP-RAM. A single 512x128 macro stores a complete pruned and quantized CNN model to achieve 98.8% inference accuracy on the MNIST data set and 89.0% on the CIFAR-10 data set, with a 573.4-giga operations per second (GOPS) peak throughput and a 49.4-tera operations per second (TOPS)/W energy efficiency.