US-TECHNOLOGY-ENGINEERING-RESEARCH-SEMICONDUCTOR
Taiwan is accelerating the development of silicon photonics and co-packaging optics technologies, and plans to conduct cross-certification with overseas laboratories in 2027 to strengthen the AI high-speed computing supply chain.

The government is stepping up efforts to develop silicon photonics by planning a verification laboratory and launching reciprocal accreditation with overseas labs in 2027. The goal is to obtain certification from the Optical Internetworking Forum (OIF) in 2028, shortening the time required to send Taiwan-developed products overseas for testing and strengthening the competitiveness of the domestic semiconductor and AI high-performance computing supply chain.

According to an Oct. 30 report by Central News Agency, National Development Council Minister Kung Ming-hsin said AI has become a key driver of Taiwan's economic growth. The government has identified silicon photonics, quantum technology and AI robotics as priority areas for future development.

In silicon photonics, the government plans to promote co-packaged optics (CPO), ultra-high-speed and low-power optical transmission technologies to meet the demand from AI servers and data centers for high-speed data transmission.

Kung said it currently takes about six months at the shortest and as long as two years for Taiwan's silicon photonics products to undergo verification at overseas laboratories. Establishing domestic verification capabilities and pursuing reciprocal accreditation with laboratories in the Netherlands, the United States, Belgium and France is expected to shorten testing and commercialization timelines.

Besides silicon photonics, Taiwan is also continuing to advance quantum technology. Kung said Taiwan completed development of a 10-qubit superconducting quantum chip this year and will launch the second phase of the program next year, expanding international cooperation and promoting the commercialization of quantum technology.

Silicon photonics uses optical signals to transmit data. Compared with conventional electrical transmission, it offers the potential for higher speeds and lower energy consumption in some transmission applications. As AI models grow larger, data centers face rising demands for high-speed interconnects and energy efficiency, making the technology a key potential direction for the next generation of computing infrastructure.

However, the technology's ability to deliver commercial competitiveness will still depend on verification standards, manufacturing costs, industry cooperation and the pace at which international customers adopt it. Whether Taiwan can establish a complete research, development, testing and mass-production ecosystem will be crucial to its future industry strategy.