An RTL code generation method, a computer device and a readable storage medium
By constructing a structure graph and a bidirectional mapping index, and combining feedback from large models and EDA tools, the RTL code generation process is automatically checked and repaired. This solves the problems of low efficiency and reliance on manual labor in existing methods, and achieves efficient and automated RTL code generation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CORE TREND (ZHUHAI) TECH CO LTD
- Filing Date
- 2026-05-06
- Publication Date
- 2026-06-05
AI Technical Summary
Existing methods for generating RTL code based on large language models suffer from problems such as implicit expression of design elements making it difficult to automate consistency verification, lack of static checks, and EDA tool feedback reports not forming a closed loop. This results in long RTL code generation times, high costs, and reliance on manual repair.
By constructing a structure diagram for compliance checks, generating initial RTL code and performing comprehensive performance checks, and using a combination of large models and rules to generate RTL code, a bidirectional mapping index between the structure diagram and the RTL code is established. EDA tool reports are automatically parsed to drive structure diagram updates, enabling automated repair and partial updates.
It improves the efficiency and automation of RTL code generation, reduces human intervention, enables early detection and fixing of design errors, and lowers iteration costs.
Smart Images

Figure CN122152323A_ABST
Abstract
Citation Information
Patent Citations
Hardware security vulnerability knowledge graph assisted Verilog security code generation method and device
CN119558238A
Method and system for automatically checking and correcting PCB (Printed Circuit Board) design rule
CN120874746A
Integrated circuit design whole-process collaborative verification and consistency guarantee method
CN121328425A
Agentic Large Language Model (LLM) Apparatus and Method for Synthesizable Register Transfer Level (RTL) Code Generation With Progressive Feedback
US20260023539A1