Encapsulated Error Signature for Functional Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current pre-silicon functional simulation techniques lack an efficient mechanism for automatically grouping and managing simulation failures, making it time-consuming to identify and address errors.
Innovation Solution
A method and system for generating an encapsulated error signature during functional simulation by receiving error descriptor notification messages, updating arrays in real-time, and creating error clusters based on error ranking, code, and count.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional techniques are used to manually group simulation failures by going through log files, then error grouping can be achieved, but the process becomes time-consuming and tedious
Solution Approach 1:
The system performs preliminary actions by automatically generating error signatures and grouping failures during the simulation process itself, rather than requiring manual analysis after simulations complete. The error signature generator continuously monitors simulation output and pre-processes failure data into organized clusters, eliminating the need for time-consuming manual log file analysis.
Solution Approach 2:
The system enables self-service by automatically triaging and grouping simulation failures without human intervention. The error signature generator autonomously analyzes simulation output, identifies failure patterns, and organizes failures into groups based on error characteristics, allowing the system to manage its own error analysis without requiring manual processing of log files.
2Reliability
If manual analysis of log files is performed to identify errors, then error identification can be achieved, but the process becomes tedious and inefficient
Solution Approach 1:
The system replaces the mechanical process of manual log file analysis with an automated computational system. The error signature generator uses algorithmic processing to analyze simulation output, extract error patterns, and group failures automatically, substituting human manual work with machine-based error identification that maintains accuracy while dramatically improving efficiency.
Solution Approach 2:
The error signature acts as an intermediary between raw simulation output and human analysis. Instead of requiring direct manual examination of extensive log files, the system generates condensed error signatures that capture essential failure information, serving as an intermediate representation that maintains identification accuracy while reducing analysis time and effort.
3Ease of operation
If no automated error grouping mechanism is implemented, then system complexity remains low, but error management becomes time-consuming and manual
Solution Approach 1:
The system applies segmentation by dividing error management into distinct functional components: error detection, signature generation, and failure grouping. This modular approach breaks down the complex task of error management into manageable segments that can be processed automatically, improving ease of operation while managing system complexity through structured organization of the error signature generation process.
Data Source
AI summary
This disclosure relates to system and method for generating an encapsulated error signature during functional simulation. The method includes receiving at least one error descriptor notification message from at least one of a plurality of testbench components. Each of the at least one error descriptor notification message includes values corresponding to a plurality of error attributes including error ranking, error code, error message, and error score. The method further includes iteratively updating in real-time, a plurality of arrays based on values corresponding to one or more of the plurality of error attributes, in response to receiving each of the at least one error descriptor notification message. The method further includes iteratively generating in real-time, an encapsulated error signature based on each of the error ranking, the error code, and an error count associated with the error code derived from one or more of the plurality of arrays.