We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Hardware Architecture

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

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.
Comments: This work has been accepted by IEEE Journal of Solid-State Circuits (JSSC 2021)
Subjects: Hardware Architecture (cs.AR); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
Journal reference: IEEE Journal of Solid-State Circuits, Volume: 56, Issue: 6, Pages: 1924 - 1935, June 2021
DOI: 10.1109/JSSC.2021.3056447
Cite as: arXiv:2107.02388 [cs.AR]
  (or arXiv:2107.02388v1 [cs.AR] for this version)

Submission history

From: Kaiyuan Yang [view email]
[v1] Tue, 6 Jul 2021 04:59:16 GMT (3848kb)

Link back to: arXiv, form interface, contact.