Revolutionary Light-Based AI Chip: Smaller Than a Speck of Dust, Faster Than Traditional Computing

Scientists create revolutionary light-based AI chip smaller than a speck of dust

TL;DR: Engineers are exploring AI to advance scientific fields, with a new AI chip using light for faster, energy-efficient calculations. Engineers are attempting to discover new ways to leverage AI to make breakthroughs in various scientific fields, and one may have just happened that harnesses the power and speed of light. Researchers have penned a new study published in Nature Photonics that builds upon an earlier concept initially unveiled in 2018, where researchers showcased the power of diffractive neural networks. For those who don't know, traditional computing is comprised of electronic circuit boards, and while extremely powerful in their own right, they do have limitations or concessions users must make. Electronic circuit boards have inherent latency as it takes time for the data to be moved. Additionally, this process can be extremely energy-demanding. However, researchers have designed a new AI chip that manipulates light as it's capable of performing the calculation instantly versus a traditional computer that has to interpret light. As light travels through the new AI chip, it's directed, significantly speeding up the transmission process of data and also reducing the power needed for a calculation to be completed. While being a ground-breaking design that has many implications for numerous fields of computing, the new light-manipulating AI chip does face significant problems, and that is scaling it at the product. What the light-based AI chip is currently producing According to the researchers, the chip is made with a fixed design, meaning it will need to be customized for an individual task. Unfortunately, engineers won't be able to just produce 100,000 of these chips and roll them out globally, they will need to be made one by one, and specifically designed for the process they will be undertaking. "As the tasks change or the fibre-optic system changes, one would need to have a new design to be fabricated and integrated with the fibre," says Aydogan Ozcan at the University of California, Los Angeles At the moment, the new AI chip can process data flowing through fiber-optic cables as fast as light travels through each of the layers, which means the computations can be performed within trillionths of a second.

This AI chip is the size of a grain of salt

Fiber optic cables have a bottleneck problem. Although they're capable of transporting encoded data at the speed of light, translating the encoded data into understandable information often requires slower, much more energy-hungry equipment. Building off of previous innovations in a field known as passive neural networks, however, a team at China's University of Shanghai for Science and Technology (USST) is developing a microscopic workaround: A new artificial intelligence chip that utilizes light physics to analyze data using only a fraction of the energy. What's more, each chip is barely the size of a grain of salt. The recent advances, detailed in a study published in the journal Nature Photonics rely on a form of neural networking first developed by researchers at the University of California, Los Angeles in 2018. Known as an "all-optical diffractive deep neural network," this method uses patterned, 3D-printed layers of passive components that are precisely stacked together. The system is then trained to complete complex computations utilizing photons of light. As New Scientist explains, the USST team recently used this concept as their starting point to create a "passive, well-trained neural network" that physically maneuvers light to perform computational analysis. All that data-encoded light, however, is conveyed through fiber optic wires no wider than a single human hair -- so they needed their AI chip to be just as tiny in order to read each photon. Researchers relied on "three-dimensional two-photon nanolithography" to construct each miniscule chip using ultrathin polymer layers. They then attached a chip to the end of a fiber optic wire, where it processed data as it passed through the cables at the speed of light. To test the invention, the team encoded images of numerals into light photons, then sent them through the fiber optic wires. The AI chips then successfully read the data and recreated each number image with minimal fuzziness. This kind of image recognition is now a rudimentary function in many AI systems, the salt-sized chips managed to do so in trillionths of a second. They also do so using only a few thousandths of the amount of energy as today's AI-based image recognition technology. The system is by no means perfect just yet. The slightest chip imperfections can degrade the overall system, and each chip must be specifically customized depending on the job needed. Still, the inventors believe the technology could eventually provide "unprecedented functionalities." These might include situations such as endoscopic imaging and potentially even for quantum computing.

AI chip smaller than a grain of salt uses light to decode data

A tiny chip on the tip of a fibre-optic cable can passively harness light to perform AI computations, dramatically reducing the amount of energy and computing power required An AI chip smaller than a grain of salt can perch at the end of an optical fibre, harnessing the physics of light to process the information passed through the fibre - while using far less energy and computational power than typical AI techniques. Optical fibres can carry data at the speed of light, but decoding these light signals usually happens on a slower and more energy-intensive external computing device. The new AI chip could perform that data-processing task more...