Ricursive

The goal of the company is to speed the transition from idea to working silicon. Their long term goals are divided into three phases:

• Phase one will use AI to speed and improve chip design, primarily physical design and verification.
• Phase two will go from workload to completed chips
• Phase three will be to co-design of workload and chips, specifically to enable fast iteration.

Although the company has stated "We are definitely not an EDA company", in phase one they are taking on precisely the tasks typically performed by these companies. Therefore their commercial competitors will include Cadence Design Systems, Synopsys, and Siemens EDA. Academic software for these tasks is combined in the OpenROAD Project.

Initial products include floorplanning, placement, routing, and verification.

AlphaChip

AlphaChip is a deep reinforcement learning method for automated chip floorplanning. The basic ideas were introduced in a 2021 paper from Google, which describes an approach to macro placement, a stage of chip floorplanning. It is based on reinforcement learning (RL), a machine learning method in which a system iteratively improves its decisions by optimizing performance-based reward signals.

The claims of the seminal AlphaChip paper have engendered considerable controversy.