A 22-nm 4.92 TOPS/W End-to-End RNS DNN Accelerator for Edge-AI Devices
- 8 May 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computing
- Vol. 14 (2) , 604-618
Abstract
This work presents an end-to-end Residue Numbering System (RNS) Deep Neural Network (DNN) accelerator targeting edge-AI devices. The proposed architecture translates the intrinsic advantages of RNS-particularly the efficient implementation of multiply-add operations-into system-level power efficiency gains. This is made possible by the introduction of novel architectural features, such as the amortization of high-complexity RNS operations (base extension, activation, and scaling), and the integration of bespoke RNS low-power techniques, such as a clock-gating scheme that exploits the periodic usage of the high-complexity RNS units, and voltage scaling that exploits the RNS capability to achieve clock frequency constraints with smaller supply voltages. Tradeoffs between hardware performance metrics of RNS implementations for various operational scenarios and RNS bases have been systematically analyzed, and comparisons against a conventional positional binary implementation have been conducted, guiding the optimal selection of design space parameters. Silicon power measurements on the 22-nm FDSOI prototype chips underpin the theoretical analysis and simulation results, which show considerable benefits of RNS-based DNN processing. These prove that RNS can increase the maximum achievable frequency of arithmetic circuits and result in 1.34× more energy-efficient processing compared to conventional binary counterparts. Compared to the state-of-the-art RNS-based DNN accelerator, the proposed architecture is 9× more power efficient, reaching a peak power efficiency of 4.92 TOPS/W.Keywords
All Related Versions
Funding Information
- Khalifa University of Science, Technology and Research
- Semiconductor Research Corporation
This publication has 36 references indexed in Scilit:
- An end-to-end RNS CNN Accelerator2024
- Improving Residue-Level Sparsity in RNS-based Neural Network Hardware Accelerators via Regularization2023
- Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference2018
- Edge computing architecture for applying AI to IoT2017
- Algorithms for comparison in residue number systems2016
- Deep Residual Learning for Image Recognition2016
- Modulo-(2^n -- 2^q -- 1) Parallel Prefix Addition via Excess-Modulo Encoding of Residues2015
- Microsoft COCO: Common Objects in Context2014
- Arithmetic Units for RNS Moduli {2n-3} and {2n+3} Operations2010
- ImageNet: A large-scale hierarchical image database2009