Founders Anna Goldie and Azalia Mirhoseini introduced a system that uses artificial intelligence to learn from each silicon iteration, aiming to deliver faster, more power‑efficient chips for data centers and emerging AR/VR workloads.

Ricursive Intelligence has unveiled an AI‑driven chip design platform that promises to accelerate the development of data‑center and AR/VR hardware.

AI‑Powered Design Loop

Founders Anna Goldie and Azalia Mirhoseini describe a system that ingests data from each silicon prototype, using machine learning to predict performance and power characteristics for the next iteration.

Benefits for Data Centers

By automating layout optimization and timing closure, the platform can shave weeks off the traditional design cycle, enabling faster rollout of chips that consume less energy while delivering higher throughput.

Targeting AR/VR Workloads

AR/VR applications demand low latency and high bandwidth; Ricursive’s AI engine tailors transistor placement and interconnect strategies to meet those constraints without sacrificing battery life.

  • Reduced design time from months to weeks
  • Improved power‑efficiency through predictive optimization
  • Scalable across multiple process nodes

The company plans to integrate the platform with major foundries, offering a cloud‑based interface that lets engineers submit design goals and receive AI‑generated layout suggestions.

For readers interested in broader AI trends, the AI category on uxc.news provides additional coverage of emerging technologies.

Read the full story on TechCrunch coverage of Ricursive Intelligence’s AI‑designed chips.

Track coverage on Momentum · More in AI.