How to Automate Device Model Parameter Extraction and Re-Centering

Get Quote
Electronic design automation
+ Electronic design automation
Electronic design automation
+ Electronic design automation

Automate Machine Learning Parameter Extraction and Model Re-Centering

Compact model parameter extraction starts by selecting the measured plots or figures of merit to fit, identifying the model parameters associated with those targets, and defining suitable parameter ranges. Continuous current-voltage (IV), pulsed IV, capacitance-voltage (CV), S-parameters, temperature scaling, and geometry scaling can be included as fitting targets. The machine learning optimizer then evaluates the selected plots and parameters together through a global optimization rather than requiring a sequence of separate fitting operations.

For model re-centering, target data is loaded into the Model Generator and scaling plots are created for the required device geometries. Engineers select the desired re-centering plots, parameters, parameter ranges, error function, sampler, iteration limit, and random seed before running the optimization. The resulting simulated curves are compared with measured data, allowing the parameter set to be optimized across the selected targets and reused as part of a repeatable extraction workflow.

Automated Device Model Parameter Extraction and Re-Centering Solution

Device model re-centering requires measured targets, corresponding model parameters, defined parameter ranges, and an optimization process that minimizes fitting error across the selected data. The IC-CAP ML Optimizer combines multiple fitting plots and parameters into a global optimization, reducing the need to divide the extraction into a sequence of small optimization steps. Model Generator supports loading target data, generating scaling plots, selecting re-centering targets, and managing the fitting process. Engineers can configure the sampler, Relative Root Mean Square Error (RMSE), maximum iterations, and random seed. The same approach can be incorporated into customized extraction flows for re-centering and more complex compact-model extraction such as BSIM4 modeling.
Get Quote

Automatic Re-centering Enabled by Machine Learning

Explore Products in Our Automated Device Model Parameter Extraction and Re-Centering Solution

Discover Resources and Insights

Additional Resources for Automated Device Model Parameter Extraction and Re-Centering

Related Use Cases

How to Optimize a PIN Photodetector Model for a Silicon Photonics Foundry Process

Tune a PIN photodetector model to measured data and verify its response across multiple bias conditions for a silicon photonics foundry process.

Learn More

How to Design Flip-Chip Transmon Resonators

Evaluate flip-chip transmon and CPW resonator behavior by comparing resonator orientation relative to adjacent silicon and metal layers.

Learn More

How to Model Superconducting Kinetic Inductance

Model kinetic inductance and resonant-frequency shifts in temperature-dependent superconducting circuits.

Learn More

contact us logo

Get in Touch with One of Our Experts

Need help finding the right solution for you?

Contact Us