ML Sequence Generation Ecosystem for IC Design Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating test sequences in integrated circuit design verification are time-consuming, prone to human error, and do not effectively cover all corner cases.
Innovation Solution
A machine learning-based sequence generation ecosystem that explores and identifies different states by dividing valid operations into actions and action sequences, using online inference to discover new states and generate test sequences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If test sequences are generated manually or randomly, then verification can be performed, but the process is time-consuming and prone to human error
Solution Approach 1:
The patent replaces manual mechanical processes of test sequence generation with machine learning-based automated generation. The ML models learn from design specifications and automatically generate comprehensive test sequences, eliminating human manual effort while improving both speed and reliability of verification processes.
Solution Approach 2:
The system enables self-service verification by having the machine learning models autonomously generate test sequences without human intervention. The models continuously learn and improve from feedback, automatically adapting to design changes and generating appropriate test cases independently, thereby reducing dependency on manual verification processes.
2Reliability
If manual test generation is used, then some test cases can be created, but corner cases are not effectively covered
Solution Approach 1:
The patent employs parameter changes by adjusting ML model parameters and exploration strategies to systematically vary test conditions. This enables comprehensive coverage of corner cases by exploring different parameter spaces that manual methods would miss, while the automated nature handles the complexity of managing these parameter variations.
Solution Approach 2:
The system implements feedback mechanisms where test results and coverage information are fed back to the ML models, which then refine their test sequence generation. This continuous feedback loop ensures improving test coverage over time, automatically identifying and testing corner cases without increasing manual process complexity.
3Productivity
If human-controlled test generation is used, then tests can be created, but errors and omissions occur
Solution Approach 1:
The patent replaces human-controlled mechanical processes with automated machine learning-based test generation. This substitution eliminates human errors and omissions while maintaining high productivity, as the ML models can generate tests faster and more accurately than manual processes without sacrificing efficiency.
Data Source
AI summary
An approach is disclosed herein a sequence generation ecosystem using machine learning. The approach disclosed herein is a new approach to sequence generation in the context of validation that relies on machine learning to explore and identify ways to achieve different states. In particular, the approach divides the valid operations into different respective actions and action sequences. These actions are selected by machine learning models as they are being trained using online inference reinforcement learning. This online inference also is likely to result in the discovery of new states. Each state that has been identified is then used as a target to train a respective machine learning model. As part of this process a representation of all the states and actions or sequences of actions executed to reach those states is created. This representation, the respective machine learning models, or a combination thereof can then be used to generate different test sequences.