Generative AI Connectivity File Generation for DUT Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The process of connecting third-party verification tools to a Design Under Test (DUT) in a verification testbench is complex and time-consuming, requiring manual coding and expert knowledge, especially in modern digital circuit designs with numerous connections.
Innovation Solution
A generative AI-based tool infrastructure that automates the generation of verification testbench components, including property modules, bind modules, and connectivity files, using a retrieval augmented generation (RAG) framework to improve the quality of large language model (LLM) responses by integrating domain-specific code syntax and user feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual coding is used to connect third-party verification tools to DUT, then expertise and control are maintained, but time consumption and complexity increase significantly
Solution Approach 1:
The system enables self-service by automatically generating connectivity files through AI that connect third-party verification tools to the DUT without requiring manual expert intervention. The AI model autonomously creates the necessary connection code based on the verification requirements.
Solution Approach 2:
Manual mechanical coding work is replaced by an AI-based automated system. The generative AI model substitutes the manual process of writing connectivity code with automated code generation, significantly reducing the time and effort required while maintaining connection accuracy.
2Manufacturing precision
If manual coding is used to create connectivity files, then precise control over connections is achieved, but the process becomes extremely complex and time-consuming
Solution Approach 1:
The complex manual process of creating connectivity files is replaced by an AI-based automated system. The generative AI model handles the complexity of generating accurate connection code automatically, reducing the perceived complexity for users while maintaining high connection accuracy.
Solution Approach 2:
The AI model acts as an intermediary between the verification engineer's requirements and the actual connectivity code generation. It translates high-level verification requirements into precise connection code, managing the complexity internally while presenting a simplified interface to users.
3Reliability
If expert knowledge is required for manual connectivity file creation, then connection quality is maintained, but productivity decreases due to the need for specialized skills
Solution Approach 1:
The system empowers verification engineers to self-generate connectivity files using AI without requiring deep expert knowledge of connection protocols and tools. The AI model encapsulates expert knowledge internally, allowing users to achieve reliable connections through automated generation rather than manual expert intervention.
Solution Approach 2:
Expert manual coding is replaced by AI-based automated code generation. The generative AI model substitutes the need for specialized expert skills with automated intelligence, maintaining connection reliability while dramatically increasing productivity by eliminating the bottleneck of expert availability.
Data Source
AI summary
Embodiments of the present disclosure include a software solution for generating a connectivity file that forms the connections between signals in a design under test (DUT) and signals in a verification IP (VIP) interface. The connecting signals from the DUT may form a DUT interface and the DUT interface may follow a protocol. The software solution may utilize a generative AI-based tool to identify example connections that also follow the protocol. These example connections may be included in a query to a large language model (LLM) for purposes of improving the results generated by the LLM.