Insider Brief
- Astrus, a Toronto- and Waterloo-based AI startup, raised $8M led by Khosla Ventures to build the world’s first physics-aware foundation model for chip design.
- Its reinforcement learning system automates the analog layout process — traditionally months of manual engineering — by generating thousands of high-quality layouts in seconds
- Founded by Brad Moon and Zeyi Wang, with a team trained under AI pioneers, Astrus will use the funding to expand research, scale compute infrastructure, and deliver tools to leading semiconductor companies.
PRESS RELEASE — Astrus, the AI startup building the world’s first physics-aware foundation model for chip design, announced it has raised $8 million USD in funding. The round was led by Khosla Ventures with participation from Juniper Networks founder Pradeep Sindhu, 1517 Fund, Drive Capital, Alumni Ventures, and other strategic investors.
While digital chip design has become increasingly automated, the analog portion of advanced semiconductors remains a painstaking, manual process. Engineers place transistors one at a time, guided by experience and intuition. These layouts can take months to produce, with costs reaching into the hundreds of millions for cutting-edge designs.
Astrus’s AI is changing that. The company’s reinforcement learning architecture learns the underlying physics of chip design, generating thousands of high-quality layouts in seconds. What once required months of manual engineering can now be achieved in hours, accelerating the development of advanced chips and enabling faster progress in compute.
“Analog layout is one of the biggest bottlenecks in the design process for the most advanced microchips,” said Brad Moon, co-founder and CEO of Astrus. “Our AI learns the physics behind chip design, allowing us to automate what has always been a manual process. This funding enables us to accelerate chip development for the world’s leading teams and drive faster progress in compute.”
Astrus was founded in Toronto and Waterloo by Brad Moon, a former satellite-sensor chip designer, and Zeyi Wang, a reinforcement learning researcher trained by AlphaGo advisor Martin Müller. Their team also includes Kenny Young, who completed his PhD under reinforcement learning pioneer Rich Sutton. By adapting the core ideas behind AlphaGo and applying them to analog layout, Astrus is building a foundation model capable of surpassing human designers and uncovering entirely new circuit architectures.
“We’re scaling toward one of the largest reinforcement learning training runs ever attempted, engineered to push chip design past human limits,” said Kenny Young, Astrus’s Founding Research Scientist.
With this new funding, Astrus will expand its research and engineering team, scale compute infrastructure for large-scale RL training, and deliver tools to leading semiconductor companies working on the most advanced designs.
About Astrus
Astrus is building the world’s first physics-aware foundation model for chip design. By automating the most complex parts of analog layout, Astrus enables faster progress in computing and pushes the boundaries of semiconductor innovation. Learn more at astrus.ai.
Contacts
Media Contact
Karson Simpson
Public Relations Manager
Communitech
karson.simpson@communitech.ca
