Calibration & Services

Synopsys TCAD tools are used by process and device engineers at all major semiconductor companies to develop and optimize semiconductor technologies. Occasionally, the tools need to be calibrated for a specific technology, so that they can be more predictive for future nodes.

Synopsys TCAD Services offers calibration, simulation, model development, and consulting to customers. Our team has extensive know-how in helping customers to deploy TCAD-based technology development methodologies, resulting in a significant increase in engineering productivity, reduction in development costs, and improvement in product yield. Drawing upon a wide spectrum of experience and expertise from TCAD modeling and technology development, our experts deliver turnkey solutions for consistent, accurate, and predictive process–device modeling of a wide range of technologies, such as Logic, Analog and Discrete, Memory, Optical, and Electrostatic Discharge (ESD).

Accelerate TCAD calibration with automated, machine learning-based workflows that help engineers achieve accurate, predictive simulation results with less setup and fewer simulations.

Expert calibration modules provide up to 80% of the calibration workflow pre-built, helping engineers quickly get started with a productivity boost in calibration workflow creation. Enhanced machine learning capabilities, including active learning, advanced sampling techniques, and uncertainty-aware neural networks, make it easier to create efficient, reusable calibration workflows.

Sentaurus ML Calibration Accelerator

Scale compute for demanding ML-based calibration workflows with Sentaurus ML Calibration Accelerator, a new companion product to Sentaurus Calibration Workbench. It works with existing calibration workflows and helps engineers rapidly recalibrate TCAD models as new test data becomes available.

• Get started faster with expert-curated calibration modules
• Extract metrology profiles from TEM and SEM data with manual or guided workflows
• Reduce DOE requirements with active learning and advanced sampling
• Create and reuse customized calibration modules and workflows
• Accelerate ML model training with uncertainty-aware neural networks

Model for elimination of lifetime-limiting carbon vacancy defects in SiC by thermal treatment (Journal of Applied Physics)