Chip Design

AI helps engineering teams design, analyze, and validate chips faster and with fewer design errors.

Business impact

• engineering productivity — The evidence describes AI as improving productivity and helping teams move faster, though one account notes the benefit had not yet been established in a large trial.
• design cycle time — Amazon says the tools help them move faster across the design cycle.
• reliability — AMD says it uses AI in chip-design work to improve reliability.

Data requirements

• chip-design materials (text) — Used to answer questions and generate technical documentation.
• bug reports (text) — Used for summarization in the NVIDIA pilot.
• engineering code and scripts (text) — Used for code completion and script generation.
• simulation and analysis data (tabular) — Used in AI-supported simulation and analysis workflows.

AI methods and techniques

• question answering — Used to answer chip-design questions for engineers.
• summarization — Used to summarize bug reports and technical material.
• code completion — Used in internal R&D to help engineers write code faster.
• document generation — Used to generate technical documentation from chip-design materials.
• agentic AI — Used to support engineering workflows across design, analysis, optimization, and validation.
• simulation and analysis support — Used with EDA and multiphysics workflows to help evaluate chip and system designs.

AI models and model families

domain-adapted LLMs, ChipNeMo

Industries

Real-world evidence

4 documented case studies on record.

Companies using this: AMD, MediaTek, NVIDIA, Amazon.