Higuang Xin Approaching HKEx Listing Hearing with 0.8% Global Optical Module Market Share
Alina Collins
Silicon-photonics optical module maker Hylight (海光芯正) has cleared its HKEX listing hearing, holding 0.8% of the global optical module market. Revenue grew nearly sixfold over three years, yet cumulative losses topped RMB 200 million — whether AI-datacenter demand can push it past breakeven is the central question for this IPO.
What does this company actually make?
Hylight's core product is optical modules — small devices that convert electrical signals into light, letting servers inside data centers talk at high speed.
Its key differentiator: end-to-end in-house capability from chip design to module manufacturing, built around silicon photonics — a technology that moves signals with light instead of electricity, delivering higher speed at lower power.
Products span 100G to 800G. Every single-mode module at 400G and above uses silicon photonics, targeting AI data centers.
Where does it rank in the industry?
By 2025 revenue, Hylight ranks 17th among global optical module suppliers, with a 0.8% market share.
This means → across the full optical-module landscape, it is still a small player.
But in the AI-optical-module sub-segment alone, it ranks 8th globally among Chinese suppliers, with share rising to 1.6%. In plain terms = the company is concentrating its chips on AI, the fastest-growing line in the sector.
What else is in the product lineup?
AOC (active optical cables): optical module plus fiber fused into one cable, used for short-distance links inside or between racks.
AEC (active electrical cables): copper cables with signal-amplification chips at each end — commercial shipments began December 2025.
Silicon-photonics AOCs (400G / 800G / PCIe 6.0) are still in development, aimed at next-generation high-bandwidth, low-power interconnects.
How are the financials?
Revenue grew nearly sixfold in three years: roughly RMB 175 million in 2023 → RMB 862 million in 2024 → RMB 1.221 billion in 2025.
Yet the company lost money every year: annual losses of roughly RMB 109 million, RMB 17.9 million, and RMB 100 million, totaling over RMB 200 million.
This reflects a classic "growing revenue, not profit" bind — top-line expansion is running ahead of the scale needed to absorb R&D and capacity costs.
What should investors watch?
The central question is straightforward: can AI-datacenter demand push this company past breakeven?
This means → two things to track: whether downstream AI-compute buildout keeps delivering orders, and whether silicon-photonics yields and unit costs improve with scale.
In plain terms = revenue growth proves the market wants what Hylight makes, but if profitability stays elusive after the IPO raise, valuation pressure will arrive quickly.
Content is for reference only, not financial advice.