Adaptive Energy—Accuracy Trade-Offs in Configurable MAC Architectures for AI Acceleration
Open Access
- 9 March 2026
- journal article
- research article
- Published by MDPI AG in Electronics
- Vol. 15 (5) , 1129
Abstract
Energy efficiency has become a primary bottleneck in hardware platforms supporting machine learning workloads, particularly as modern inference and training tasks demand sustained high-throughput computation. This challenge is further amplified in energy-harvesting and intermittently powered systems, where the available energy budget varies over time. This work introduces a run-time configurable multiply–accumulate (MAC) architecture that dynamically adjusts arithmetic precision to match instantaneous energy availability. The proposed design relies on an internally adaptive multiplier based on bit-level logic compression, enabling controlled modulation of power consumption while preserving numerical robustness. Crucially, the MAC maintains a fixed external operand interface, allowing for seamless precision adaptation without operand reformulation or datapath disruption. The architecture is implemented in System Verilog and evaluated using both ASIC synthesis in a 90 nm CMOS technology and FPGA deployment. Experimental results demonstrate approximately a fourfold improvement in power–delay product (PDP) relative to full-precision operation, with only limited degradation in inference accuracy.Keywords
Funding Information
- Taibah University, Madinah, Kingdom of Saudi Arabia (1126-15-447)
This publication has 28 references indexed in Scilit:
- Acceleration Techniques for Automated Design of Approximate Convolutional Neural NetworksIEEE Journal on Emerging and Selected Topics in Circuits and Systems, 2023
- High-Performance Accurate and Approximate Multipliers for FPGA-Based Hardware AcceleratorsIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2021
- Neural network design for energy-autonomous artificial intelligence applications using temporal encodingPhilosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2019
- Input-Conscious Approximate Multiply-Accumulate (MAC) Unit for Energy-EfficiencyIEEE Access, 2019
- Design and Analysis of Area and Power Efficient Approximate Booth MultipliersIEEE Transactions on Computers, 2019
- Real-Power ComputingIEEE Transactions on Computers, 2018
- Lifetime improvement by exploiting aggressive voltage scaling during runtime of error-resilient applicationsIntegration, 2018
- Bio-Inspired Imprecise Computational Blocks for Efficient VLSI Implementation of Soft-Computing ApplicationsIEEE Transactions on Circuits and Systems I: Regular Papers, 2009
- Energy Harvesting From Human and Machine Motion for Wireless Electronic DevicesProceedings of the IEEE, 2008
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998