Silicon photonics has recently become a hot topic due to its higher transmission efficiency compared with electrical signals, making it a critical component for future artificial intelligence (AI) computing. According to Wealth Magazine Biweekly, experts believe that beyond AI and high-performance computing (HPC), silicon photonics is expected to expand into more application areas, including network communications, biomedical sensing, and autonomous driving, which are attracting significant attention.

The Industrial Technology Research Institute (ITRI) under Taiwan’s Ministry of Economic Affairs notes that silicon photonics integrates electronic and optical components using mature silicon wafers and semiconductor processes, effectively increasing data transmission speeds on chips.

U.S. Leads Silicon Photonics Research, Growth Accelerates Post-2021

Reviewing silicon photonics research, the United States benefits from early investment and diverse applications, holding a global lead. In 2014, it established the Integrated Photonics Manufacturing Institute, promoting the industrialization and practical application of the technology.

According to Lin Chuan-Jun, analyst at the Second Research Division of the Chung-Hua Institution for Economic Research, influenced by AI industry development, the number of National Science Foundation (NSF) research projects related to silicon photonics has grown significantly since 2021. Recent funding focuses on quantum photonics and data center optical interconnects.

With rising demand in AI and HPC, many NSF projects prioritize using silicon photonics to accelerate data processing and reduce computational energy consumption.

Lin emphasized that silicon photonics applications will not be limited to AI and HPC. “Network communications, biomedical sensing, and autonomous driving (optical LiDAR) are future focal points.” Among these three, she believes network communications is likely to see earlier adoption and offers a more advantageous niche—essentially becoming “optical communication.”

Biomedical sensing can leverage mature semiconductor processes, but its widespread application, such as in cancer detection, is limited compared to network communications, which everyone uses daily, generating larger market demand. Similarly, applying silicon photonics in autonomous driving LiDAR can improve self-driving technology, but the market adoption rate is lower than that of network communications.

Europe Focuses on Applications, Taiwan Equipment Makers Poised to Benefit

“Applications like biomedical sensing, optical LiDAR, or environmental monitoring won’t see the same scale as network communications or HPC because it’s tied to product volume and demand,” Lin said.

Wealth Magazine Biweekly analyzed that Europe is actively investing in applied silicon photonics. Lin noted that European companies may not focus heavily on advanced semiconductor processes but are enhancing silicon photonics deployment in network communications, biomedical sensing, and autonomous driving.

For instance, in biomedical sensing, the University of Southampton in the U.K. is participating in a cancer detection research project aimed at developing low-cost, high-performance, highly sensitive silicon photonics mid-infrared (MIR) sensing chips, enabling mass production and driving biomedical applications.

Given that network communications may see early adoption of silicon photonics beyond AI and HPC, the question arises: can Taiwan’s industry benefit? Lin shared insights from industry forums, noting that Taiwan has many network equipment manufacturers. If silicon photonics drives further development in network communications, creating related equipment demand, Taiwanese companies could gain early entry into this emerging technology market.

Resource: 矽光子不會只用在AI！未來先攻網路通訊　台廠設備商迎潛在商機