GenEDA introduces the first aligned encoder-decoder foundation model for electronic design automation, unifying circuit representation and generation capabilities.

• Integrates both netlist analysis and circuit generation in a single foundation model
• Creates a shared latent space between encoder and decoder for seamless multimodal operations
• Demonstrates superior performance in circuit understanding, reasoning, and generation tasks
• Enables practical applications like netlist repair and functionality recovery

This breakthrough could significantly accelerate IC design workflows by leveraging AI to understand and generate circuit designs, potentially reducing development cycles for complex electronic systems.

GenEDA: Unleashing Generative Reasoning on Netlist via Multimodal Encoder-Decoder Aligned Foundation Model