Can light outperform electricity in AI computing? A new research lab aims to unlock faster processing with dramatically lower power consumption.
Researchers at Shanghai Jiao Tong University have launched the Shanghai Key Laboratory of Integrated Photonic Computing Chips and Systems, a dedicated research facility focused on developing photonic computing technologies for future AI hardware.
The laboratory, described as China’s first industry academia platform dedicated to integrated photonic computing, will explore how light based processors can address the growing power and performance challenges facing artificial intelligence systems.
As AI models become larger and more complex, conventional semiconductor technologies are approaching physical and energy efficiency limits. Data centres supporting AI workloads require increasing amounts of computing power, creating demand for alternatives that can deliver higher performance without proportionally increasing power consumption.
Photonic computing is emerging as one such alternative. Unlike traditional chips that move data using electrons through silicon circuits, photonic chips use photons, or light particles, for data transmission and computation. Because light travels faster and generates significantly less heat, photonic systems have the potential to offer higher bandwidth, lower latency, and improved energy efficiency.
The laboratory will focus on several core areas including photonic chip architectures, silicon photonics integration, optical components, algorithms, and practical commercial applications. Researchers aim to address the scientific and engineering challenges that currently prevent photonic computing from achieving its theoretical performance advantages.
According to Zou Weiwen, director of the laboratory and professor at Shanghai Jiao Tong University, photonic computing represents “An important pathway for achieving breakthroughs in computing power,” citing advantages in bandwidth, latency, and energy efficiency.
The launch reflects growing global interest in next generation computing technologies as researchers seek ways to sustain AI growth beyond the limits of conventional semiconductor scaling.
