Intern_Experimentation of Large Language Models for Mixed-Signal Verification - Bevaix, Switzerland
JOB.TITLE
-
OU.DIVISION
-
EMPLOYMENT.TYPE
-
OU.LOCATION
Your future mission
Modern Integrated Circuits have more features and safety mechanisms, expanding the verification scope, while shorter time-to-market pressures drive the need for automation. As a Melexis intern, you will work with the mixed-signal verification team to explore AI-driven automation for verification tasks. This can be a Master's thesis or an internship.
More specifically, you will:
Monitor technological developments in AI/LLM/RAG topics.
Train a RAG-based large language model on an existing codebase of verification components & test cases.
Use the RAG-based LLM to generate new components based on verification specifications.
Compare the performance of various models and frameworks.
Integrate the developed tools into the verification flow.
Expected results
- State of the art of the technology and available tools
- Proof of concept of RAG-based LLM able to generate basic verification components
- Documentation
Requirements
Interest in latest Artificial Intelligence developments. Hands-on experience with custom AI tools is a plus
Prior experience with Python or other scripting language is a plus
Experience with SystemVerilog is an asset