Duke ECE Hardware Startup Selected for Y Combinator

A hardware startup spun out of Duke’s Department of Electrical and Computer Engineering (ECE) has been selected for Y Combinator’s upcoming summer batch, placing it among a select group of companies invited to join one of Silicon Valley’s most influential accelerator programs.

The startup, Synapse Semiconductor, was founded by Tania Roy, associate professor of ECE, and recent Duke graduate Sanjeev Chauhan (BS ’26, MS ’26). Y Combinator, which has helped launch companies like Airbnb, DoorDash and OpenAI, accepts less than 0.05% of applicants each cycle. Hardware companies, and particularly semiconductor companies, are even rarer.

The selection offers an external validation of the innovation ecosystem at Duke Engineering and an example of how emerging technology can be translated to a global stage.

Retina-Inspired Hardware

Synapse Semiconductor is building what they describe as a “retina on a chip,” a device that mimics the human eye by processing visual information at the moment it is captured.

In most AI and computer-vision systems, camera sensors passively capture all visual data and send it to separate processors, which then extract the features that matter. For example, a self-driving car collects an entire scene—a nearby car slowing down, a stop light ahead, falling leaves, sun glare, a plume of dust, etc.—and only after that data reaches a processor can the system determine what’s relevant. The human retina, by contrast, filters noise, enhances edges and identifies key features before sending the information to the brain.

Roy’s lab has been working for several years to emulate that biological efficiency in hardware. Their devices use ultrathin materials such as molybdenum sulfide to create pixels that can gather and process data directly at the sensor.

“We’re trying to build a pixel that doesn’t just sense light but can think about that light,” Roy said. “It processes information where it is captured, just like the retina.”

This approach reduces lag, energy and memory overhead. Already, they’ve attracted interest from potential partners, especially companies developing robots, drones or satellites that operate in energy-constrained environments or lack access to cloud computing.

Translating Research at Duke ECE

Roy and Chauhan noted you could count the number of hardware companies in a year at Y Combinator on one hand.

“Most companies at Y Combinator are software,” Roy said. “We showed up with working hardware, not just a pitch. Pixels that only sense light are outdated. Ours sense and think in the same device, and we’ve already built and measured them.”

Chauhan joined Roy’s lab as an undergraduate researcher and continued as a master’s student. Along the way, he built experience on the business side through Duke Capital Partners, giving him a clearer view of how university research becomes a company and how to raise venture capital.  

“There are so many entrepreneurial resources—Duke Capital Partners, the Innovation & Entrepreneurship programs, Office for Translation & Commercialization, New Ventures, the CFCI accelerator,” Chauhan said. “All in support of spinning out research to commercialization.”

Roy said the lab’s rapid progress since she arrived at Duke three years ago is due to the strong graduate students and a culture at Duke Engineering and Duke ECE that encourages new ideas.

“This isn’t just about our lab. It shows that our department has the talent, resources and ecosystem to turn semiconductor research into real-world hardware,” Roy said. “If we can do it, others can too.”