|
Functional coverage is a growing area of interest within the verification of complex hardware designs, as it is the most effective means for monitoring the quality of verification and directing the stimuli generators to areas that have not been adequately verified. The various activities in the Machine Learning and Constraint Satisfaction Group are focused on functional coverage and its relations to other components in the simulation-based verification flow, mainly test-generation.
The group is responsible for the development of support for a cross-product functional coverage methodology. In this context, we develop and maintain a trace-driven functional coverage tool - Meteor. This tool is developed to facilitate and support our functional coverage methodology. Meteor supports user-defined coverage models at various levels of abstraction: architecture, micro-architecture, and design implementation.
One of the main bottlenecks of the verification process in general, and coverage analysis in particular, is closing the loop between the coverage results and directives to the stimuli generators. To address this bottleneck, we research and develop coverage-directed generation (CDG) methodology and technology, which is designed to automate the process of using feedback from coverage analysis for tuning generation stimuli towards areas not adequately verified. CDG casts the problem as a statistical inference problem, and uses Bayesian Networks to encode the complex joint input-ouput distribution space, for ultimately inferring generation directives to the stimuli generator.
| 
