MIT scientists have developed a new photonic chip that can optically perform all the key computations of a deep neural network.

The deep neural network models that power today’s most demanding machine-learning applications have grown so large and complex that they are pushing the limits of traditional electronic computing hardware.

Photonic hardware, which can perform machine-learning computations with light, offers a faster and more energy-efficient alternative. However, there are some types of neural network computations that a photonic device can’t perform, requiring the use of off-chip electronics or other techniques that hamper speed and efficiency.

The optical device was able to complete the key computations for a machine-learning classification task in less than half a nanosecond while achieving more than 92 percent accuracy — performance that is on par with traditional hardware.

The chip, composed of interconnected modules that form an optical neural network, is fabricated using commercial foundry processes, which could enable the scaling of the technology and its integration into electronics.

In the long run, the photonic processor could lead to faster and more energy-efficient deep learning for computationally demanding applications like lidar, scientific research in astronomy and particle physics, or high-speed telecommunications.

“There are a lot of cases where how well the model performs isn’t the only thing that matters, but also how fast you can get an answer. Now that we have an end-to-end system that can run a neural network in optics, at a nanosecond time scale, we can start thinking at a higher level about applications and algorithms,” says Saumil Bandyopadhyay ’17, MEng ’18, PhD ’23, a visiting scientist in the Quantum Photonics and AI Group within the Research Laboratory of Electronics (RLE) and a postdoc at NTT Research, Inc., who is the lead author of a paper on the new chip.

Machine learning with light

Deep neural networks are composed of many interconnected layers of nodes, or neurons, that operate on input data to produce an output. One key operation in a deep neural network involves the use of linear algebra to perform matrix multiplication, which transforms data as it is passed from layer to layer.

However, in addition to these linear operations, deep neural networks perform nonlinear operations that help the model learn more intricate patterns. Nonlinear operations, like activation functions, give deep neural networks the power to solve complex problems.

In 2017, Englund’s group, along with researchers in the lab of Marin Soljačić, the Cecil and Ida Green Professor of Physics, demonstrated an optical neural network on a single photonic chip that could perform matrix multiplication with light.

But at the time, the device couldn’t perform nonlinear operations on the chip. Optical data had to be converted into electrical signals and sent to a digital processor to perform nonlinear operations.

They overcame that challenge by designing devices called nonlinear optical function units (NOFUs), which combine electronics and optics to implement nonlinear operations on the chip.

The researchers built an optical deep neural network on a photonic chip using three layers of devices that perform linear and nonlinear operations.

Integrated networks

At the outset, their system encodes the parameters of a deep neural network into light. Then, an array of programmable beamsplitters, which was demonstrated in the 2017 paper, performs matrix multiplication on those inputs.

The data is then passed to programmable NOFUs, which implement nonlinear functions by siphoning off a small amount of light to photodiodes that convert optical signals to electric current. This process, which eliminates the need for an external amplifier, consumes very little energy.

Achieving such low latency enabled them to efficiently train a deep neural network on the chip, a process known as in situ training that typically consumes a huge amount of energy in digital hardware.

MIT Technology Review