Our ability to cram ever-smaller transistors onto a chip has enabled today’s age of ubiquitous computing. But that approach is finally running into limits, with some experts declaring an end to Moore’s Law and a related principle, often known as Dennard’s Scaling.

Those developments couldn’t be coming at a worse time. Demand for computing power has skyrocketed in recent times thanks largely to the rise of artificial intelligence, and it shows no signs of slowing down.

Now Lightmatter, an organization founded by three MIT alumni, is constant the remarkable progress of computing by rethinking the lifeblood of the chip. As an alternative of relying solely on electricity, the corporate also uses light for data processing and transport. The corporate’s first two products, a chip specializing in artificial intelligence operations and an interconnect that facilitates data transfer between chips, use each photons and electrons to drive more efficient operations.

“The 2 problems we’re solving are ‘How do chips talk?’ and ‘How do you do these [AI] calculations?’” Lightmatter co-founder and CEO Nicholas Harris PhD ’17 says. “With our first two products, Envise and Passage, we’re addressing each of those questions.”

In a nod to the dimensions of the issue and the demand for AI, Lightmatter raised just north of $300 million in 2023 at a valuation of $1.2 billion. Now the corporate is demonstrating its technology with a few of the largest technology firms on the earth in hopes of reducing the large energy demand of knowledge centers and AI models.

“We’re going to enable platforms on top of our interconnect technology which can be made up of a whole lot of 1000’s of next-generation compute units,” Harris says. “That simply wouldn’t be possible without the technology that we’re constructing.”

From idea to $100K

Prior to MIT, Harris worked on the semiconductor company Micron Technology, where he studied the elemental devices behind integrated chips. The experience made him see how the standard approach for improving computer performance — cramming more transistors onto each chip — was hitting its limits.

“I saw how the roadmap for computing was slowing, and I desired to work out how I could proceed it,” Harris says. “What approaches can augment computers? Quantum computing and photonics were two of those pathways.”

Harris got here to MIT to work on photonic quantum computing for his PhD under Dirk Englund, an associate professor within the Department of Electrical Engineering and Computer Science. As a part of that work, he built silicon-based integrated photonic chips that might send and process information using light as a substitute of electricity.

The work led to dozens of patents and greater than 80 research papers in prestigious journals like . But one other technology also caught Harris’s attention at MIT.

“I remember walking down the hall and seeing students just piling out of those auditorium-sized classrooms, watching relayed live videos of lectures to see professors teach deep learning,” Harris recalls, referring to the factitious intelligence technique. “Everybody on campus knew that deep learning was going to be an enormous deal, so I began learning more about it, and we realized that the systems I used to be constructing for photonic quantum computing could actually be leveraged to do deep learning.”

Harris had planned to grow to be a professor after his PhD, but he realized he could attract more funding and innovate more quickly through a startup, so he teamed up with Darius Bunandar PhD ’18, who was also studying in Englund’s lab, and Thomas Graham MBA ’18. The co-founders successfully launched into the startup world by winning the 2017 MIT $100K Entrepreneurship Competition.

Seeing the sunshine

Lightmatter’s Envise chip takes the a part of computing that electrons do well, like memory, and combines it with what light does well, like performing the large matrix multiplications of deep-learning models.

“With photonics, you possibly can perform multiple calculations at the identical time because the info is coming in on different colours of sunshine,” Harris explains. “In a single color, you may have a photograph of a dog. In one other color, you may have a photograph of a cat. In one other color, perhaps a tree, and you may have all three of those operations going through the identical optical computing unit, this matrix accelerator, at the identical time. That drives up operations per area, and it reuses the hardware that is there, driving up energy efficiency.”

Passage takes advantage of sunshine’s latency and bandwidth benefits to link processors in a fashion much like how fiber optic cables use light to send data over long distances. It also enables chips as big as entire wafers to act as a single processor. Sending information between chips is central to running the large server farms that power cloud computing and run AI systems like ChatGPT.

Each products are designed to bring energy efficiencies to computing, which Harris says are needed to maintain up with rising demand without bringing huge increases in power consumption.

“By 2040, some predict that around 80 percent of all energy usage on the planet shall be dedicated to data centers and computing, and AI goes to be an enormous fraction of that,” Harris says. “Whenever you have a look at computing deployments for training these large AI models, they’re headed toward using a whole lot of megawatts. Their power usage is on the dimensions of cities.”

Lightmatter is currently working with chipmakers and cloud service providers for mass deployment. Harris notes that because the corporate’s equipment runs on silicon, it may possibly be produced by existing semiconductor fabrication facilities without massive changes in process.

The ambitious plans are designed to open up a latest path forward for computing that might have huge implications for the environment and economy.

“We’re going to proceed taking a look at all the pieces of computers to work out where light can speed up them, make them more energy efficient, and faster, and we’re going to proceed to switch those parts,” Harris says. “Without delay, we’re focused on interconnect with Passage and on compute with Envise. But over time, we’re going to construct out the subsequent generation of computers, and it’s all going to be centered around light.”