Researchers at Cornell University have developed a low-power chip they call a “microwave brain.” It’s the first processor that can handle ultra-high-speed data and wireless signals using microwaves.
Researchers at Cornell University have developed a low-power chip they call a “microwave brain.” It’s the first processor that can handle ultra-high-speed data and wireless signals using microwaves.
The processor is the first true microwave neural network and is fully integrated into a silicon chip. It performs real-time frequency-domain computations for tasks such as decoding radio signals, tracking radar targets, and processing digital data, while consuming less than 200 milliwatts of power, Tech Xplore reports .
"Because the chip can programmatically change a signal over a wide range of frequencies in a fraction of a second, it can be used for a variety of computing tasks," explained Bal Govind, the study's lead author. "It avoids many of the signal processing steps typically performed by digital computers."
This capability is provided by the chip's design as a neural network, a computer system modeled after the brain, using interconnected modes created in tunable waveguides.
This allows it to recognize patterns and learn from data. But unlike traditional neural networks, which rely on digital operations and step-by-step instructions synchronized to a clock, this network uses analog, nonlinear behavior in the microwave regime, which allows it to process data streams at tens of gigahertz — much faster than most digital chips.
The chip can perform both low-level logic functions and complex tasks, such as identifying bit sequences or counting binary values in high-speed data. It has achieved 88% or better accuracy on several classification tasks involving types of wireless signals, comparable to digital neural networks but with a fraction of the power and size.
"In traditional digital systems, as tasks become more complex, more circuitry, more power, and more error correction are required to maintain accuracy. But with our probabilistic approach, we can maintain high accuracy in both simple and complex calculations without additional overhead," the researchers say.
The chip's extreme sensitivity to input signals makes it suitable for hardware security applications, such as detecting anomalies in wireless communications across multiple microwave frequency bands.
Furthermore, the researchers believe that if they reduce power consumption even further, they could deploy it on a smartwatch or mobile phone.
Although 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.
