MIT's light-speed AI chip foresees the era of 6G high-speed networking

2025/06/27

TechNews Technology News
June 20, 2025
Author Cindy An

With the rapid increase in connected devices, applications such as video conferencing, cloud gaming and smart homes are increasingly demanding bandwidth, and managing limited wireless spectrum has become a major challenge. The key to solving this problem lies in high-performance real-time processing technology, and the latest breakthrough by a research team at the Massachusetts Institute of Technology (MIT) is expected to completely change the status quo.

The optical AI processor developed by MIT, which has been published in the journal Science Advances, can process wireless signals at the speed of light, achieving unprecedented computing efficiency. This chip, called MAFT-ONN (Multiplicative Analog Frequency Conversion Optical Neural Network), can complete signal classification in nanoseconds, which is nearly 100 times faster than existing digital systems and consumes much less energy than traditional AI hardware.

Unlike traditional digital processors that need to convert signals into images before processing, MAFT-ONN operates directly in the frequency domain, skipping the digital conversion step, achieving significant energy saving and acceleration. This chip can achieve 85% classification accuracy in one measurement, and more than 99% in multiple measurements. The whole process takes only 120 nanoseconds, far better than existing technologies.

MAFT-ONN has a unique design. Each neural network layer only needs one device. It can integrate 10,000 neurons on a single chip and use "photoelectric multiplication" technology to achieve high-efficiency computing. The research team must develop its own machine learning architecture to match the hardware characteristics to ensure that the physical characteristics of the chip can be fully utilized.

This technology is particularly important in the future 6G network, enabling smart devices such as "cognitive radio" to adapt to network conditions in real time and maintain a stable connection. Ronald Davis III, a member of the research team, said that as the measurement time increases, the accuracy will increase, and the ultra-fast processing time of MAFT-ONN makes this process completely unaffected by performance.

Dirk Englund, a professor in the Department of Electrical and Computer Science at MIT, pointed out: "This technology opens up many possibilities for real-time and reliable AI inferences, and is the beginning of far-reaching impact." This chip can not only change wireless communications, but also be applied to fields such as real-time environmental response of self-driving cars and smart medical devices.

The research team plans to apply "reuse technology" to improve chip performance and expand to more complex deep learning architectures, including Transformer models and large language models. This breakthrough will have a profound impact on various technical fields that require real-time and efficient AI, laying the foundation for the era of true intelligent connectivity.

【Disclaimer】
The content of this article only represents the author's personal views and has nothing to do with Creating.
The content, text description and originality have not been verified by this site. This site does not make any guarantee or commitment to this article and all or part of the content, authenticity, completeness, and timeliness. It is for readers' reference only. Please verify the relevant content yourself.

Creating Nano Technologies, Inc.
59 Alley 21 Lane 279, Chung Cheng Road, Yung Kang City, Tainan, TAIWAN
TEL:886-6-2323927 FAX:886-6-2013306 URL: http://www.creating-nanotech.com

BACK TO LIST