![[2510.13582] ArtNet: Hierarchical Clustering-Based Artificial Netlist Generator for ML and DTCO Application](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2510.13582] ArtNet: Hierarchical Clustering-Based Artificial Netlist Generator for ML and DTCO Application
Summary
ArtNet introduces a novel artificial netlist generator that enhances machine learning model generalization and design-technology co-optimization by producing realistic datasets.
Why It Matters
As the complexity of optimizing power, performance, and area (PPA) in advanced nodes increases, ArtNet addresses the challenges of limited training data and long design turnaround times. This innovation is crucial for improving machine learning applications in hardware design, making it relevant for researchers and engineers in the field.
Key Takeaways
- ArtNet enhances ML model generalization through realistic dataset generation.
- The artificial netlist generator supports broader design space exploration for DTCO.
- ArtNet improves F1 scores in CNN-based DRV prediction by 0.16.
- Generated mini-brains achieve a PPA match of up to 97.94%.
- The approach addresses challenges in power, performance, and area optimization.
Computer Science > Machine Learning arXiv:2510.13582 (cs) [Submitted on 15 Oct 2025 (v1), last revised 19 Feb 2026 (this version, v2)] Title:ArtNet: Hierarchical Clustering-Based Artificial Netlist Generator for ML and DTCO Application Authors:Andrew B. Kahng. Seokhyeong Kang, Seonghyeon Park, Dooseok Yoon View a PDF of the paper titled ArtNet: Hierarchical Clustering-Based Artificial Netlist Generator for ML and DTCO Application, by Andrew B. Kahng. Seokhyeong Kang and 2 other authors View PDF HTML (experimental) Abstract:In advanced nodes, optimization of power, performance and area (PPA) has become highly complex and challenging. Machine learning (ML) and design-technology co-optimization (DTCO) provide promising mitigations, but face limitations due to a lack of diverse training data as well as long design flow turnaround times (TAT). We propose ArtNet, a novel artificial netlist generator designed to tackle these issues. Unlike previous methods, ArtNet replicates key topological characteristics, enhancing ML model generalization and supporting broader design space exploration for DTCO. By producing realistic artificial datasets that moreclosely match given target parameters, ArtNet enables more efficient PPAoptimization and exploration of flows and design enablements. In the context of CNN-based DRV prediction, ArtNet's data augmentationimproves F1 score by 0.16 compared to using only the original (real) dataset. In the DTCO context, ArtNet-generated mini-brains achie...
