AdAM: Adaptive Fault-Tolerant Approximate Multiplier for Edge DNN Accelerators
Mahdi Taheri, Natalia Cherezova, Samira Nazari, Ahsan Rafiq, Ali Azarpeyvand, Tara Ghasempouri, Masoud Daneshtalab, Jaan Raik, Maksim Jenihhin
arXiv:2403.02936v116 citationsh-index: 20ETS
Originality Synthesis-oriented
AI Analysis
This work addresses hardware reliability issues for edge computing applications, but appears incremental as it builds on existing approximate multiplier techniques.
The paper tackled the problem of designing a fault-tolerant approximate multiplier for ASIC-based DNN accelerators, proposing an adaptive architecture to improve reliability and efficiency.
In this paper, we propose an architecture of a novel adaptive fault-tolerant approximate multiplier tailored for ASIC-based DNN accelerators.