China has inaugurated its first dedicated photonic computing laboratory at Shanghai Jiao Tong University, marking a strategic move to develop light-based AI chips amid increasing US export curbs on conventional semiconductors.
- Lab partnership unites Shanghai Jiao Tong University with photonic startup Lightelligence
- Photonic chips use light particles, promising faster, more efficient AI processing
- The program aligns with China's $295 billion plan to develop domestic AI infrastructure
What happened
On June 11, China opened its first dedicated photonic computing lab in Shanghai, hosted at Shanghai Jiao Tong University and operated jointly with Lightelligence, a leading Chinese photonic computing startup. The lab will focus on research into photonic chip architectures, silicon-photonics integration, and the development of optical components and algorithms essential for commercial viability.
Lightelligence, known for its claim to pioneer large-scale hybrid optical-electronic computing deployments, recently went public on the Hong Kong stock exchange, underscoring both the technological and financial momentum behind China's push into photonic computing.
Why it matters
Photonic processors differ from traditional AI chips by using photons instead of electrons to transmit data, which can enable higher bandwidth, lower latency, and significantly reduced heat and energy consumption. These attributes are crucial as the computational demands of artificial intelligence accelerate beyond the capabilities of conventional silicon chips.
With the US imposing strict export controls on advanced semiconductor technology to China since 2022, Beijing views photonic computing as a promising alternative path to bypass these limits. By leveraging existing strengths in fiber optics and laser technologies, China hopes to reduce its reliance on foreign semiconductor supply chains while pushing forward on AI chip innovation.
What to watch next
Despite the significant potential, photonic computing remains an emerging technology facing substantial scientific and engineering challenges, including the need for a mature software and algorithm ecosystem to effectively utilize optical hardware. Progress at the new lab in Shanghai will be critical to moving from laboratory concepts to scalable, commercial AI chip products.
China's broader AI ambitions include a massive $295 billion plan to build domestic data centers running primarily on homegrown chips by 2028. The success of photonic computing initiatives like this lab could determine how China navigates global semiconductor supply challenges over the coming decade.