Advertisement
Lightelligence on track with IPO plans as China’s AI photonics race gathers pace
Shanghai-based company aims to become Hong Kong’s first listed AI photonics chipmaker as AI infrastructure buildout drives momentum
2-MIN READ2-MIN
1
Listen
Minxiao Changin Shenzhen
Silicon photonic computing chips – long overlooked in the artificial intelligence hardware stack – are emerging as a new focal point in mainland China’s semiconductor push, as domestic companies move towards public listings amid intensifying US-China competition and surging demand for next-generation computing infrastructure.
Shanghai-based Lightelligence, the first company globally to achieve large-scale deployment of hybrid optical-electronic computing, passed its Hong Kong listing hearing on Monday, advancing plans for an initial public offering (IPO) expected to raise between US$300 million and US$400 million. That puts the company on track to become the first AI photonics chipmaker to list in Hong Kong.
The listing is set to be one of the largest in the city this year, testing investor sentiment amid US-China tech tensions, though Lightelligence said its business was not subject to US export control restrictions.
Advertisement
The rapid buildout of hyperscale AI clusters is fuelling demand for high-speed, low-latency interconnect technologies, with research firm Frost & Sullivan forecasting the global AI computing and interconnect market to grow at a 27 per cent compound annual rate to 2031.
Optical computing, once seen as niche, is gaining relevance as AI systems scale, with photons used to transmit and process information instead of traditional electronic pathways, helping to reduce latency, bandwidth and power constraints in data centres. Optical interconnect uses photonics to enable high-speed, low-latency data transfer between chips and systems such as graphics processing units (GPUs) and central processing units.
Lightelligence focuses on hybrid optical-electronic computing, with operations spanning optical interconnects and optical computing. The company described its technology as a complement to GPUs rather than a replacement, targeting bottlenecks in data transmission and matrix computation within AI systems.
Advertisement
Advertisement
Select Voice
Select Speed
1.00x