Search
Menu

Photonic Computing System Rethinks How to Power Familiar AI Tool

Facebook X LinkedIn Email
Researchers from MIT have demonstrated a photonics-based computing system that could lead to machine-learning programs several orders of magnitude more powerful than the one behind ChatGPT. The researchers reported a greater than 100-fold improvement in energy efficiency and a 25-fold improvement in compute density — a measure of system power — over state-of-the-art digital computers for machine learning.

In the near term, the researchers’ experimental, laser-based system could be further developed to improve these metrics by two more orders of magnitude, the researchers said.

According to research team member Dirk Englund, an associate professor in MIT’s Department of Electrical Engineering and Computer Science, ChatGPT is limited in its size by the power of modern supercomputers. Deep neural networks (DNNs) like the one behind ChatGPT are based on large machine-learning models that simulate how the brain processes information. However, the digital technologies behind DNNs are reaching their limits even as the field of machine learning is growing. They require massive amounts of energy and are largely confined to data centers. It is also not economically viable to train models that are much bigger, Englund said. 

The newly demonstrated system uses hundreds of micron-scale VCSEL arrays to perform computations based on the movement of light rather than electrons. Though optical neural networks (ONNs) typically use a great deal of energy, the researchers believe they could advance their system to cut down on
energy usage. At the same time, they could offer the potential for larger bandwidths, said researcher and lead author Zaijun Chen.

The researchers said that by supporting the development of large-scale optoelectronic processors to accelerate machine-learning tasks from data centers to decentralized edge devices, cellphones, and other small devices could become capable of running programs that can currently only be computed at large data centers. 
They acknowledged that the components involved in ONNs are bulky and take up significant space. And, although ONNs are quite good at linear calculations such as adding, they are not great at nonlinear calculations such as  multiplication and “if” statements.


Still, the researchers expect that the approach could be scaled to commercial use in a few years. The components of the system can be created using fabrication processes already in use. Chen said that the laser arrays used for the system are widely used today in cellphone face ID and data communication.  

Artist's rendition of a computer system based on light that could jumpstart the power of machine-learning programs like ChatGPT. Blue sections represent the micron-scale lasers key to the technology. Courtesy of Ella Maru Studio.


Artist’s rendition of a computer system based on light that could jump-start the power of machine-learning programs like ChatGPT. Blue sections represent the micron-scale lasers key to the technology. Courtesy of Ella Maru Studio. 
The researchers have filed for a patent on the work, which was sponsored by the Army Research Office, NTT Research, the National Defense Science and Engineering Graduate Fellowship Program, the National Science Foundation, the Natural Sciences and Engineering Research Council of Canada, and the Volkswagen Foundation.

The research was published in Nature Photonics (www.doi.org/10.1038/s41566-023-01233-w).

Published: August 2023
Glossary
quantum
The term quantum refers to the fundamental unit or discrete amount of a physical quantity involved in interactions at the atomic and subatomic scales. It originates from quantum theory, a branch of physics that emerged in the early 20th century to explain phenomena observed on very small scales, where classical physics fails to provide accurate explanations. In the context of quantum theory, several key concepts are associated with the term quantum: Quantum mechanics: This is the branch of...
neural network
A computing paradigm that attempts to process information in a manner similar to that of the brain; it differs from artificial intelligence in that it relies not on pre-programming but on the acquisition and evolution of interconnections between nodes. These computational models have shown extensive usage in applications that involve pattern recognition as well as machine learning as the interconnections between nodes continue to compute updated values from previous inputs.
machine learning
Machine learning (ML) is a subset of artificial intelligence (AI) that focuses on the development of algorithms and statistical models that enable computers to improve their performance on a specific task through experience or training. Instead of being explicitly programmed to perform a task, a machine learning system learns from data and examples. The primary goal of machine learning is to develop models that can generalize patterns from data and make predictions or decisions without being...
Research & TechnologyLasersOpticscomputingquantumneural networkmachine learningChatGPTlanguage modelMITDirk EnglundRyan Hamerlyoptical neural networkVCSELvertical cavity surface emitting laserNature PhotonicsTechnology News

We use cookies to improve user experience and analyze our website traffic as stated in our Privacy Policy. By using this website, you agree to the use of cookies unless you have disabled them.