Newsgather
BackPhotonics Emerges as Next Computing Frontier Amid AI-Driven Demand
Photonics Emerges as Next Computing Frontier Amid AI-Driven Demand
Tech
SCMP Economy5/31/2026Tech2 min readChina

Photonics Emerges as Next Computing Frontier Amid AI-Driven Demand

Scientists and Investors Turn to Light-Based Technology to Overcome Copper Interconnect Limitations in Data Centers

Quick Look

As AI computing demands outpace traditional copper interconnect capabilities, photonics (light-based tech) gains investment and attention for its lower-loss optical links, with China's Yuanjie Technology seeing an 11x stock surge.

AI-generated summary

Why It Matters

The scaling of AI computing clusters has reached physical limits with traditional copper interconnects, driving the need for photonics.

Font size

As artificial intelligence strains the physical limits of existing data centres, scientists and investors are turning to the ultimate speed limit of the universe for the next computing frontier: light. For Mi Lei, founder of CAS Star, a venture capital firm born out of a research institute under the state-run Chinese Academy of Sciences (CAS), the sudden global fascination with photonics is less a surprise than a delayed validation. It is a thesis he has spent more than a decade trying to support with funding. “New technologies can drive industrial upgrading,” Mi said from his Shanghai office, which carries a nameplate “Wallfacer”, after the elite strategists in award-winning sci-fi author Liu Cixin’s The Three-Body Problem trilogy whose minds serve as humanity’s last refuge from alien surveillance. “Our role is to push scientific research towards commercialisation.” Light matters Mi, who holds a doctorate in optics from the Xian Institute of Optics and Precision Mechanics under CAS, has been investing in photonics from day one. Today, more than 200 of CAS Star’s roughly 600 portfolio companies span the broader photonics sector, covering sensing, communications, computing, storage and display. Optics have been drastically revalued as AI clusters scale to unprecedented sizes. To train huge AI models, tens of thousands of specialised computing chips must constantly share data with one another. Traditionally, this data is transmitted through standard copper cables and wires. However, these copper interconnects are now hitting a hard physical wall. Shoving large amounts of data through electricity causes signal loss, generates excessive heat and skyrockets power consumption – creating a bottleneck that slows down the chips. This has made lower-loss optical links, which use light instead of electricity, central to modern data-centre design. China’s stock investors are aware. Yuanjie Technology, a Shaanxi-based laser-chipmaker that CAS Star invested in around 2019, has seen its shares surge more than elevenfold over the past year, fuelled by surging revenue from data-centre light-source products.

What to Watch

AI outlook — possibilities, not facts

  • Increased global investment in photonics research and development

    Very likely · Within months

Open Questions

  • How will global competition in photonics evolve?
  • What are the potential environmental impacts of widespread photonics adoption?

Related Topics

This article was originally published by SCMP Economy.

Related Stories

More on this topicphotonics