At the University of Pennsylvania, engineers have developed a chip enabling algorithms to operate using light pulses instead of traditional circuits and silicon.
This is the globe's premier photonic processor capable of training artificial intelligence systems in real time with the use of light beams. Such innovation has the potential to revolutionize the way we develop AI, which plays an ever-growing role in our daily existence.
Training artificial intelligence consumes a lot of energy. This chip turns the tables.
Over the past few years, AI has transitioned from being a "someday" technology to an integral part of our daily lives. Underneath this transformation lies the power of neural networks—vast interconnected systems resembling artificial neurons fine-tuned through extensive training processes. While these neural networks are highly resilient, they require immense computational resources for their operation. electricity-hungry hardware .
Today, most AI runs on specialized chips called GPUs . These chips are fast but energy intensive. The cost of training cutting-edge AI models like GPT-4 can run into millions of dollars — and even more in carbon emissions .
Nonlinear functions are essential for training. deep neural networks,” says Liang Feng, the paper’s senior author, referring to the training phase. “Our aim was to make this happen in photonics for the first time.”
This is precisely where the latest chip from Penn Engineering comes into play. It utilizes light for training neural networks completely, significantly reducing energy consumption and markedly boosting processing speed.
It goes beyond efficiency; it represents a revolutionary shift: an entirely new approach to computing.
How does this work?
To grasp this advancement, let's discuss how AI "processes information."
Fundamentally, many artificial intelligence systems nowadays rely on neural networks. These networks consist of nodes (similar to neurons) interconnected with weighted links. As data moves through this structure, some connections strengthen the signal, others weaken it, and specific paths become activated.
But there’s a twist: these systems don’t just rely on adding and multiplying numbers. The real magic happens in the nonlinear functions — mathematical operations where small inputs can lead to big changes. Nonlinearity is what gives AI its power to detect patterns, recognize faces, or drive cars.
For decades, engineers dreamed of using photons instead of electrons to compute. Light is fast and doesn’t heat up like electricity. It can also travel in parallel beams, handling multiple signals at once.
But light has a problem. It travels in straight lines, and in most materials, it behaves in a linear way. That means it’s great for adding things up — but terrible for the nonlinear twists AI needs. While many teams developed light-power chips capable of handling linear mathematics, none managed to use light for non-linear functions. Until now.
The secret lies in a special semiconductor
The team’s innovation starts with a special semiconductor that reacts to light. Think of it as a thin film that can become more or less transparent, depending on how you shine light into it.
Next, the team employs two beams: one transports the information known as “signal” light, while the second serves as an unseen guide called “pump” light, shaping the way the material reacts. Through meticulous adjustment of the form, strength, and layout of the pump light, they have command over whether the signal light gets absorbed, boosted, or modified. This interplay mirrors nonlinear mathematical operations, which are utilized within neural networks for making choices.
And it’s all done without changing the chip’s physical structure.
“We’re not changing the chip’s structure,” says Feng. “We’re using light itself to create patterns inside the material, which then reshapes how the light moves through it.”
A fresh approach to artificial intelligence
This kind of design is particularly well-suited for machine learning applications. Since the chip has the capability to emulate various nonlinear behaviors as needed, it can adjust its performance during the training process. It’s this feature that renders it programmable—not only at set up but also when operational.
The team tested it on classic machine learning challenges, like distinguishing species of iris flowers or recognizing spoken words. The chip trained itself, adjusting its internal light patterns to improve accuracy over time.
For the iris dataset, it achieved 96.7% accuracy — outperforming comparable linear photonic systems . When tasked with identifying spoken words like “Bird” and “Tree,” it reached over 91% accuracy using far fewer connections than traditional digital networks. The team also proved that they can achieve major simplification in hardware and massive savings in energy.
The technology is still in its early stages. The chip uses sophisticated optical setups and precise holographic light patterns to control behavior. Scaling this up will take engineering and manufacturing breakthroughs.
But the core concept — programming light to compute nonlinearly — is now proven.
Next steps include integrating the chip with existing silicon photonic platforms , increasing the number of inputs and outputs, and exploring real-world applications in vision, speech, and robotics.
You might not notice it today, but this could be the moment we remember as the dawn of light-trained AI. It won’t just be faster, it will be fundamentally different.
The study was published in Nature Photonics .
This story originally appeared on ZME Science . Looking to become smarter each day? Subscribe to our newsletter and keep up-to-date with the most recent scientific developments.