SRAM Based Digital Custom Compute Engine for Improved Area Efficiency of AI Hardware

Narendra Singh Dhakad, Santosh Kumar Vishvakarma · 2026

Read SRAM Based Digital Custom Compute Engine for Improved Area Efficiency of AI Hardware by Narendra Singh Dhakad, Santosh Kumar Vishvakarma (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-16161.txt.

Description

Dhakad and Vishvakarma present an SRAM-based digital compute engine for AI hardware, built around a 10T SRAM cell for XNOR-based in-memory computing. Placing a full adder between in-memory multiplication cells cuts routing complexity by half. The design performs multiply-accumulate operations for binary neural networks, with a 14T adder building an N-bit ripple carry adder. Overall area efficiency improves 2.67-fold over the state of the art, with lower latency and area overhead.

Catalog record

CREATORNarendra Singh Dhakad, Santosh Kumar Vishvakarma
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE18446
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.16161

Browse the full library · About the Tower of Babel