Physics-Informed Neural Networks for Device and Circuit Modeling: A Case Study of NeuroSPICE
Abstract
We present NeuroSPICE, a physics-informed neural network (PINN) framework for device and circuit simulation. Unlike conventional SPICE, which relies on time-discretized numerical solvers, NeuroSPICE leverages PINNs to solve circuit differential-algebraic equations (DAEs) by minimizing the residual of the equations through backpropagation. It models device and circuit waveforms using analytical equations in time domain with exact temporal derivatives. While PINNs do not outperform SPICE in speed or accuracy during training, they can potentially be used as surrogate models for design optimization and inverse problems due to the differentiable feature. NeuroSPICE’s flexibility enables the simulation of emerging devices, including highly nonlinear systems such as ferroelectric memories.
Fields of Study & Subjects
Model Reduction and Neural Networks
VLSI and FPGA Design Techniques
Neural Networks and Reservoir Computing
Publication details
- Journal / source
- IEEE Electron Device Letters
- Publisher
- Institute of Electrical and Electronics Engineers (IEEE)
- Volume
- 47
- Issue
- 5
- Pages
- 1021–1024
- Publication date
- 27 February 2026