Circuit Property Generation from Signal Traces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual generation of properties for circuit designs is tedious, error-prone, and requires significant time and processing power, especially as circuit designs evolve, necessitating a more efficient method to expedite property creation and verification.
Innovation Solution
A property generation tool that automatically generates properties from signal traces by creating signal signatures, applying detection rules to identify candidate signals, and producing properties such as handshake and fairness properties, thereby reducing the number of signals that need detailed analysis and processing power.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual property creation is used, then properties can be customized based on designer knowledge, but the process is tedious and time-consuming
Solution Approach 1:
The system performs self-service by automatically generating properties from circuit design data and signal traces without requiring manual intervention. The property generation engine autonomously analyzes the circuit design, identifies relevant signals, and creates verification properties, eliminating the tedious manual coding process while maintaining reliability through systematic analysis methods.
Solution Approach 2:
The manual mechanical process of property creation is replaced with an automated computational system. The property generation engine uses algorithms and machine learning models to substitute the human designer's manual analysis and coding work, significantly reducing time while maintaining or improving property quality through consistent application of verification rules.
2Reliability
If all signals are analyzed in detail when generating properties, then comprehensive coverage is achieved, but processing power is consumed
Solution Approach 1:
The signal analysis process is segmented into multiple stages: initial filtering to identify candidate signals, detailed analysis of only those candidates, and hierarchical processing where signals are divided into groups based on their relevance to circuit functionality. This segmentation allows comprehensive coverage of important signals while avoiding unnecessary analysis of irrelevant signals, thus reducing processing power consumption.
Solution Approach 2:
Different levels of analysis quality are applied to different signals based on their importance. Critical signals that directly affect circuit behavior receive detailed analysis, while less important signals receive simplified or no analysis. This local quality approach ensures comprehensive coverage of essential properties while minimizing processing power usage on non-critical signals.
3Adaptability or versatility
If circuit design complexity increases, then functionality improves, but verification difficulty increases
Solution Approach 1:
The property generation engine is designed as a universal system that can handle various types of circuit designs regardless of complexity. It uses multi-functional analysis techniques that work across different circuit architectures, signal types, and design styles, allowing the same tool to effectively verify simple and complex circuits without requiring separate specialized methods for each case.
Solution Approach 2:
The verification approach is made dynamic by adapting the analysis depth and methods based on the detected complexity of the circuit design. The system dynamically adjusts its behavior, applying more sophisticated analysis techniques to complex circuits while using simpler methods for straightforward designs, thus managing verification complexity proportionally to the actual circuit complexity.
Data Source
AI summary
A property generation tool that automatically generates a property for a circuit design from a signal trace of the circuit design. The property generation tool receives a trace of a circuit design. The trace includes signal values for a number of signals of the circuit design over a number of clock cycles. Signal signatures are generated from one or more characteristics of the signal values. Sets of candidate signals are identified from the circuit design signals based on the signal signatures. One or more properties of the circuit design are generated based on the signal values associated with the sets of candidate signals. The property can be output, for example, for display to a user of the property generation tool. Examples of properties that are generated by the property generation tool include handshaking properties and fairness properties.