{"id":30255,"date":"2025-10-05T16:23:01","date_gmt":"2025-10-05T16:23:01","guid":{"rendered":"https:\/\/metaverseplanet.net\/blog\/?p=30255"},"modified":"2026-01-05T09:01:55","modified_gmt":"2026-01-05T09:01:55","slug":"new-revolutionary-light-powered-chip-both-accelerates-ai-and-saves-energy","status":"publish","type":"post","link":"https:\/\/metaverseplanet.net\/blog\/new-revolutionary-light-powered-chip-both-accelerates-ai-and-saves-energy\/","title":{"rendered":"New Revolutionary Light-Powered Chip Both Accelerates AI and Saves Energy"},"content":{"rendered":"\n

A new photonic chip<\/strong> powered by light, using Fresnel lenses<\/strong> and integrated lasers<\/strong>, performs AI computations<\/strong> faster and significantly reduces energy consumption<\/strong>, leaving traditional chips behind.<\/p>\n\n\n\n

Convolutional Neural Networks (CNNs)<\/strong>, one of the most critical components of artificial intelligence<\/strong>, excel in tasks like facial recognition in photos, handwriting recognition, and language translation. These networks scan raw data with small filters to highlight meaningful patterns. However, this intensive processing translates into high energy consumption<\/strong>. A large portion of CNNs’ energy use is spent on the complex calculations performed on every single pixel.<\/p>\n\n\n\n

As AI systems<\/strong> grow larger and demand more energy, this traditional method creates a significant burden on data centers<\/strong>. The continually increasing energy requirement, and thus rising costs, raises concerns that innovation<\/strong> might slow down.<\/p>\n\n\n\n

\n\n\n\n

A Light-Powered Solution to AI’s Energy Problem<\/h2>\n\n\n\n
\"\"<\/figure>\n\n\n\n

To solve this problem, researchers at the University of Florida<\/strong> have achieved a groundbreaking study. Their newly developed chip, called pJTC<\/strong> (photonic joint transform correlator<\/strong>), performs energy-intensive operations using light<\/strong> instead of electricity.<\/p>\n\n\n\n

What makes the pJTC special is its revolutionary approach to speed<\/strong> and efficiency<\/strong>. Instead of the liquid crystals or micro-mirrors used in traditional technologies, the chip can program data and filters at GHz speeds<\/strong>. Furthermore, the chip utilizes reliable photonic components<\/strong> used in optical transmitters and incorporates FT Fresnel lenses<\/strong> based on on-chip silicon photonics<\/strong> to perform complex light-based mathematical operations<\/strong> directly on the chip.<\/p>\n\n\n\n

Another innovation of the chip is its ability to perform multiple calculations simultaneously using the spectral multiplexing<\/strong> technique, thanks to its integrated lasers<\/strong>. In prototype tests, the chip was able to recognize handwritten digits with 98% accuracy<\/strong>, demonstrating performance comparable to traditional electronic processors.<\/p>\n\n\n\n

The chip’s working principle differs from conventional computing. Machine learning<\/strong> data is converted into light<\/strong>, passed through the Fresnel lenses<\/strong> which bend the light, and the complex mathematical operations are performed. When the process is complete, the light is converted back into a digital signal<\/strong>, completing the AI task<\/strong>.<\/p>\n\n\n\n

Hangbo Yang, co-author of the study at the University of Florida, explained the innovation: \u201cThis is the first time we\u2019ve put this kind of optical computing<\/strong> onto a chip and applied it to an AI neural network<\/strong>. We can pass multiple wavelengths, or colors, of light through the lenses at the same time. That\u2019s the key advantage of photonics<\/strong>.\u201d This advantage is also evident in its efficiency<\/strong> and performance<\/strong>. It can achieve 305 trillion operations per watt<\/strong> and 40.2 trillion operations per square millimeter<\/strong>. This means it could power AI across all fields, from edge devices<\/strong> to cloud services<\/strong>.<\/p>\n\n\n\n

Volker J. Sorger, the study’s lead researcher, emphasized that the new technology could be easy to integrate, noting that chip manufacturers (such as Nvidia<\/a><\/em><\/strong>) already use optical elements in some AI systems.<\/p>\n\n\n\n

You Might Also Like;<\/h3>\n\n\n