CircuitsDNA is introduced, an evolutionary framework that automatically evolves accuracy-configurable arithmetic circuits supporting multiple accuracy modes within a single circuit that eliminates all search stalls observed in conventional methods across 8/12/16-bit multipliers, while adaptive mutation provides up to 1.33 times faster convergence than its non-adaptive counterpart.
Abstract
Emerging edge AI workloads increasingly require arithmetic units that can trade computational accuracy for efficiency on demand. However, existing approximate arithmetic circuits are typically fixed-accuracy or rely on predefined structures for runtime configurability. This work introduces CircuitsDNA, an evolutionary framework that automatically evolves accuracy-configurable arithmetic circuits supporting multiple accuracy modes within a single circuit. It integrates three key features: 1) multi-threshold verifiability miter to enforce mode-specific accuracy requirements, 2) resource-limited verifiability-driven search to reduce verification overhead without sacrificing correctness, enabling efficient exploration of large circuit design, and 3) feedback-driven adaptive mutation to prioritize effective structural modifications and accelerate search convergence. Experimental results show that the 8-bit multiplier variants synthesized in 28-nm CMOS reduce the area-power product by up to 56% on INT8 DNN workload and 93% under exhaustive activity, compared with an exact 8-bit multiplier. Across CNNs and DeiTs, the accuracy loss relative to FP32 remains below 2% after fine-tuning under worst-case error (WCE) budgets of at most 1%. CircuitsDNA eliminates all search stalls observed in conventional methods across 8/12/16-bit multipliers, while adaptive mutation provides up to 1.33 times faster convergence than its non-adaptive counterpart.
A novel variation of high-effort logic synthesis called transduction is presented, which performs transformation and reduction using don’t-cares to restructure the circuit.
Yukio Miyasaka, Alan Mishchenko, John Wawrzynek et al.· 0 citations
Graph neural networks (GNNs) have shown strong potential for gate-level hardware Trojan detection, yet scalability remains a fundamental challenge. As integrated circuits grow in scale, existing GNN-based methods suffer from limited scalability and reduced discriminative power due to inherent constraints of message pas...
Ju-Yang Ye, Qi-Lin Chen, Yao-Hua Wang et al.· International Test Conferenc...· 0 citations
Stochastic computing (SC) offers a hardware-efficient alternative to binary arithmetic by encoding data as probability bit streams processed with simple logic gates. However, SC accuracy is highly sensitive to correlation between bit streams generated by Linear Feedback Shift Registers (LFSRs), particularly in cascaded...
P. Krishnapriya, M. Vinodhini, S. Ko· IEEE Access· 0 citations
FINNAS is presented, a FINN-guided hardware-aware evolutionary neural architecture search framework that jointly searches quantised MLP depth, width, and global precision settings, and ranks candidates using proxy validation accuracy together with FINN-estimated LUT usage and latency under a fully parallel mapping.
Eva Chauffour, Chang-Hong Li, G. Floros et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.