Reproduce “2D-ThermAl: Physics-Informed Framework for Thermal Analysis of Circuits using Generative AI” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Soumyadeep Chandra, Sayeed Shafayet Chowdhury, Kaushik Roy
ThermAl predicts evolving chip-temperature maps from thermal-map inputs, using a neural network trained with a heat-physics regularizer. In tests on synthetic 256×256 maps spanning 25–55°C, it reports average pixel-wise RMSE of 0.71°C and inference about 200 times faster than the compared solver. Evidence is primarily synthetic; the 2D system still needs real-world validation.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.