Cornell’s Tiny “Microwave Brain” Chip Could Transform Computing and AI

Scientists have built a “microwave brain” chip that processes information at radar-like speeds while sipping power. It could revolutionize how AI and communication devices operate, from smartwatches to satellites. Credit: Shutterstock

Cornell engineers have created the world’s first “microwave brain” — a revolutionary microchip that computes with microwaves instead of traditional digital circuits.

This tiny, low-power processor performs real-time tasks like signal decoding, radar tracking, and data analysis while consuming less than 200 milliwatts.

Cornell’s “Microwave Brain” Breakthrough

Cornell University scientists have created a new kind of low-power microchip called a “microwave brain,” capable of processing both ultrafast data and wireless communication signals by using the unique properties of microwaves.

Recently described in the journal Nature Electronics, this processor is the first fully functional microwave neural network built directly on a silicon chip. It performs real-time computations in the frequency domain for demanding tasks such as radio signal decoding, radar tracking, and digital data processing, all while consuming under 200 milliwatts of power.

“Because it’s able to distort in a programmable way across a wide band of frequencies instantaneously, it can be repurposed for several computing tasks,” said lead author Bal Govind, a doctoral student who conducted the research with Maxwell Anderson, also a doctoral student. “It bypasses a large number of signal processing steps that digital computers normally have to do.”

The chip’s performance comes from its architecture, which functions as a neural network—a system inspired by the human brain. It uses interconnected electromagnetic modes within tunable waveguides to recognize patterns and adapt to incoming information. Unlike standard neural networks that rely on digital operations and clock-timed instructions, this system operates in the analog microwave range, enabling it to process data streams in the tens of gigahertz, far exceeding the speed of most digital processors.

Throwing Out the Digital Playbook

“Bal threw away a lot of conventional circuit design to achieve this,” said Alyssa Apsel, professor of engineering, who was co-senior author with Peter McMahon, associate professor of applied and engineering physics. “Instead of trying to mimic the structure of digital neural networks exactly, he created something that looks more like a controlled mush of frequency behaviors that can ultimately give you high-performance computation.”

The result is a chip that can handle both simple logic operations and more advanced tasks, such as recognizing binary sequences or identifying patterns in high-speed data. It achieved accuracy rates of 88% or higher across several wireless signal classification challenges, matching the performance of digital neural networks while using only a fraction of their energy and space.

Smarter Computing With Less Power

“In traditional digital systems, as tasks get more complex, you need more circuitry, more power, and more error correction to maintain accuracy,” Govind said. “But with our probabilistic approach, we’re able to maintain high accuracy on both simple and complex computations, without that added overhead.”

The chip’s extreme sensitivity to inputs makes it well-suited for hardware security applications like sensing anomalies in wireless communications across multiple bands of microwave frequencies, according to the researchers.

Toward On-Device AI and Edge Computing

“We also think that if we reduce the power consumption more, we can deploy it to applications like edge computing,” Apsel said, “You could deploy it on a smartwatch or a cellphone and build native models on your smart device instead of having to depend on a cloud server for everything.”

Though the chip is still experimental, the researchers are optimistic about its scalability. They are experimenting with ways to improve its accuracy and integrate it into existing microwave and digital processing platforms.

Reference: “An integrated microwave neural network for broadband computation and communication” by Bala Govind, Maxwell G. Anderson, Fan O. Wu, Peter L. McMahon and Alyssa Apsel, 14 August 2025, Nature Electronics.

DOI: 10.1038/s41928-025-01422-1

The work emerged from an exploratory effort within a larger project supported by the Defense Advanced Research Projects Agency and the Cornell NanoScale Science and Technology Facility, which is funded in part by the National Science Foundation.