In a groundbreaking development that could transform the future of artificial intelligence, two European research teams have successfully harnessed the power of light to create ultra-fast AI computing systems using ordinary glass fibers.
The collaborative research, led by postdoctoral researchers Dr. Mathilde Hary from Tampere University in Finland and Dr. Andrei Ermolaev from the Université Marie et Louis Pasteur in France, demonstrates how intense laser pulses traveling through thin glass fibers can mimic neural network operations at unprecedented speeds.
"Instead of using conventional electronics and algorithms, computation is achieved by taking advantage of the nonlinear interaction between intense light pulses and the glass," explained Hary and Ermolaev. Their system implements a particular class of computing architecture known as an Extreme Learning Machine, inspired by neural networks.
The researchers achieved remarkable results, with test accuracies exceeding 91% in image recognition tasks while operating at speeds measured in femtoseconds—millionths of a billionth of a second. This represents processing thousands of times faster than today's electronic systems.
The breakthrough comes at a critical time as traditional electronics approach their limits in terms of bandwidth, data throughput, and power consumption. With AI models growing increasingly complex and energy-hungry, the industry faces significant challenges in scaling current technologies.
"Our models show how dispersion, nonlinearity and even quantum noise influence performance, providing critical knowledge for designing the next generation of hybrid optical-electronic AI systems," noted Ermolaev. The research team aims to eventually build on-chip optical systems that can operate in real-time outside laboratory settings.
The implications extend far beyond academic research. Potential applications range from real-time signal processing to environmental monitoring and high-speed AI inference. As data centers struggle with the enormous power demands of modern AI systems, photonic computing offers a promising path toward more sustainable and dramatically faster artificial intelligence.
The project, funded by the Research Council of Finland, the French National Research Agency, and the European Research Council, represents a significant step toward practical optical computing—a field that has seen nearly $3.6 billion in investment over the past five years as companies race to develop alternatives to traditional silicon-based systems.