12.03.2025

A groundbreaking advancement in chip technology has emerged from a collaborative effort led by researchers at Peking University, with contributions from the Chinese Academy of Sciences and the University of California, Santa Barbara. This innovation centers around a 100-gigahertz (GHz) chip that uses light instead of electrical signals to synchronize processors, paving the way for transformative advancements in artificial intelligence (AI), next-generation telecommunications, and remote sensing.

The breakthrough lies in the use of photonics—leveraging light particles (photons)—to generate clock signals, which are essential for synchronizing the internal operations of computer processors. Traditionally, processors rely on electronic oscillators to produce these clock signals, but this method is power-hungry, generates excessive heat, and struggles to achieve significant speed improvements. Moreover, electronic systems often operate at fixed clock speeds, necessitating entirely new manufacturing processes for different applications, which inflates costs.

In contrast, the new all-optical chip design redefines how clock signals are created. By utilizing an "on-chip microcomb," the team has developed a mechanism that synthesizes single-frequency and wideband signals across a broad frequency spectrum. These signals serve as reference clocks for the system's electronics. The microcomb operates by channeling light through a ring-shaped structure resembling a racetrack on the chip. As photons race around this loop at the speed of light, each lap—completed in mere billionths of a second—establishes the timing standard for the chip’s clock.

This approach not only achieves ultra-high-speed synchronization but also offers unprecedented flexibility. According to Chang Lin, the study’s lead author and an assistant professor at Peking University’s Institute of Information and Communication Technology, the chip can support multiple microwave frequency bands, making it adaptable to evolving communication standards. For instance, the same chip could facilitate everything from current 5G networks to future 6G technologies without requiring hardware upgrades—a game-changer for the telecommunications industry.

One of the standout features of the photon-based clock is its ability to surpass traditional processing speeds. While modern graphics processing units (GPUs) and central processing units (CPUs) typically operate at clock speeds of 2 to 3 GHz, the new chip boasts a staggering 100 GHz capability. Such a leap in performance enables faster execution of computational tasks, providing a substantial boost to AI development and other data-intensive applications.

Beyond sheer speed, the chip promises enhanced efficiency and reduced energy consumption. In mobile phone base stations, for example, the technology could lower operational costs while improving signal quality. Similarly, in autonomous driving systems, it could enhance perception accuracy and reaction times, critical factors for safety and reliability.

The research team has demonstrated the feasibility of mass-producing these chips, fabricating thousands of them on standard 8-inch wafers. However, challenges remain, particularly regarding stability and packaging optimization. Addressing these issues will be crucial before the technology can transition into consumer-friendly products. Despite these hurdles, the potential impact is immense.

Peking University envisions this innovation revolutionizing industries reliant on high-speed computing and advanced communications. From enabling seamless integration of emerging wireless technologies to powering smarter AI-driven devices, the photon clock represents a paradigm shift in how we approach processor design and functionality. With further refinements, this technology could redefine the boundaries of what’s possible in computing and connectivity, ushering in a new era of technological progress.