Improved transistor modeling method for cryogenic SkyWater chips

Surrogate-Assisted Compact-Model Extraction for SKY130 CMOS at 77 K

Emerging TechnologiesMachine Learning

Summary

Transistor behavior changes at very cold temperatures, making it hard to design circuits that work well in those conditions. The authors created a faster way to match a common transistor model to real measurements of SkyWater 130 nm transistors at 77 K using a neural network that mimics detailed simulations. This method helps find the best model settings more quickly and accurately, improving how well simulations match actual transistor performance. They showed that searching more parameters can reduce average errors, but sometimes simpler searches work better on certain data.

What this means in practice

  • For circuit designers: Design and simulate cryogenic circuits more accurately by using improved transistor models of SKY130 devices at 77 K.
  • For hardware test engineers: Validate low-temperature transistor models faster during chip characterization by leveraging surrogate-assisted parameter tuning.

Authors

Aaron Wang, Olivia Gzamouranis, Vaibhav Lohia, Hershel Fei Xun Weiner, Olivia Seidel, Richard Cavanaugh, Giuseppe Di Guglielmo, Nhan Tran, Davide Braga

Abstract

Cryogenic circuit design requires compact transistor models that reproduce measured low-temperature behavior and run in standard circuit simulators. We fit BSIM4 parameters to measured I-V curves of SkyWater 130 nm (SKY130) transistors at 77 K using a neural surrogate of NGSpice. The surrogate maps compact-model parameters and device geometry to the complete set of output and transfer characteristics. Extraction is a bounded least-squares search: a bank of NGSpice simulations supplies initial candidates, finite differences of the surrogate supply the local Jacobian, and NGSpice re-simulates the resulting cards before selection. On 18 measured geometries, we compare a search over seven card parameters with a search over 43 parameters selected by NGSpice sensitivity. Relative to NGSpice re-simulations of the previously released cards, the two searches reduce the in-sample mean relative root-mean-square error (RRMS) by 17% and 26%, respectively. The 43-parameter search has the lower overall error, but the seven-parameter search has the lower mean on high-bias curves, and some individual curves have higher error than with the released cards under both searches.