AI-Driven Semiconductor Device Modeling Engineer

GlobalFoundries inc.; Town of Malta (NY); Sur place; USD 69 000 - 120 000; Plein temps

Résumé du poste

GlobalFoundries seeks an AI-enabled device modeling engineer to develop and deploy ML methods that fuse device physics, silicon data, and TCAD results to accelerate model extraction and improve predictive accuracy. You will create scalable modeling workflows for CMOS, cryogenic, photonic, superconducting, and other emerging devices.

The role emphasizes reproducible Python-based workflows, collaboration with device, characterization, PDK and circuit-design teams, and clear documentation for

Qualifications

• 3–5+ years in ML, scientific computing, semiconductor device modeling
• Strong Python and ML libraries experience
• Knowledge of semiconductor device physics and transistor behavior
• Experience building reproducible data workflows
• Ability to collaborate across AI, device, modeling teams

Responsabilités

• Develop, validate ML methods for device modeling
• Build reproducible data pipelines from device data, TCAD and metadata
• Train, validate, and compare models with baselines and uncertainty analysis
• Document datasets, models, training results and usage for traceability

Connaissances

Python

NumPy

Pandas

SciPy

scikit-learn

PyTorch

TensorFlow

JAX

Machine learning

Gaussian processes

Physics-informed ML

Formation

Bachelor's or Master's in Electrical/Computer Engineering, CS, Data Science, or related field

Outils

Git

MLflow

Weights & Biases

Containers

Description du poste

GlobalFoundries seeks an AI-enabled device modeling engineer to develop and deploy ML methods that fuse device physics, silicon data, and TCAD results to accelerate model extraction and improve predictive accuracy. You will create scalable modeling workflows for CMOS, cryogenic, photonic, superconducting, and other emerging devices.

The role emphasizes reproducible Python-based workflows, collaboration with device, characterization, PDK and circuit-design teams, and clear documentation for