Print version of this Book (PDF file)
Getting Started
The IC-CAP Statistics Package helps circuit designers, device engineers, and process engineers improve device and IC yields and design more robust products. IC-CAP Statistics provides you with the tools needed to identify and analyze the inter-relationships between device model parameters and electrical test data.
Both parametric and non-parametric analysis can be performed with IC-CAP Statistics. Parametric analysis features include principal component analysis, factor analysis and multiple linear regression. With these techniques, it is easy to select the best model parameters to track in electrical test or to build models that predict SPICE parameters from dominant parameters or independent factors.
The IC-CAP Statistics Package automatically generates model files from corner models or Monte Carlo analysis. In addition, an exclusive, new method for generating worst-case model candidates, called Non-Parametric Boundary Modeling is included.
Major Benefits
The IC-CAP Statistics Package:
| • | Minimizes the number of circuit simulations required to create robust designs |
| • | Relates model parameters to manufacturing and process data |
| • | Provides realistic worst-case models for arbitrary joint-probability densities |
Major Features
The IC-CAP Statistics Package provides:
| • | Data management capability for handling large sets of extracted parameters |
| • | Multivariate statistical analysis for data reduction and the generation of worst-case corner models, parametric boundary models, or Monte Carlo analysis |
| • | Proprietary Agilent EEsof non-parametric analysis algorithms for identification of nominal models and worst-case-candidate models from arbitrary joint probability densities |
| • | Statistical plotting for viewing data and determining relationships that exist between various parameters |
| • | Statistical models that can be used within the IC-CAP environment for simulation and validation, or be imported directly into external SPICE or other simulators to perform worst-case analysis or design centering on large circuits incorporating the modeled device |
| • | Flexibility to automate procedures using IC-CAP's PEL language. |
IC-CAP Statistics can bridge the gap between manufacturing and design by assisting you in selecting an optimal set of parameters to be tracked from manufacturing through electrical test. Using factor analysis, a subset of model parameters for a particular model (e.g., BSIM3) can be identified as being responsible for the majority of the statistical variation in the device. Regression-based equations can be generated by the program, which relate the dependent model parameters to this subset of "dominant" parameters.
Thus, a small set of parameters tracked in manufacturing can serve both as a means of process monitoring and a means of generating predictive models for use in circuit simulation.
User Interface
The IC-CAP Statistics user interface is similar to the main IC-CAP program. For an explanation of graphical user interface features such as toolbars and drop-down menus, refer to Chapter 2, "Program Basics," in the User's Guide.