The Zhitong Finance App learned that Guoxin Securities released a research report saying that in the global optical module market, domestic manufacturers are already in a leading position. According to LightCounting data, among the top ten optical module manufacturers in 2022, the revenue of Chinese manufacturers has surpassed Western competitors. In the AI era, due to the accelerated iteration of optical modules, leading manufacturers with leading positions have advantages in R&D capabilities, scale advantages, supply chain capabilities, customer stickiness, and production capacity layout, and are expected to benefit the core. At the same time, changes in new technology are also having an impact on changes in the landscape, focusing on manufacturers with forward-looking layouts in fields such as LPO and silicon optics.

Investment suggestions: Continue to focus on recommending leading domestic optical device and optical module companies such as Zhongji Xuchuang (300308.SZ), Tianfu Communications (300394.SZ), Xinyisheng (300502.SZ), etc., focusing on new technology directions such as silicon light and thin-film lithium niobate, such as Yuanjie Technology (688498.SH), optical library technology (300620.SZ), etc., and it is recommended to focus on Guangxun Technology (002281.SZ).

Guoxin Securities's views are as follows:

AI drives network transformation, and high-speed optical modules benefit.

Historically, the growth in data traffic has driven optical modules to the current production value of 10 billion US dollars. At this stage, AI has become an elastic source of data traffic growth, and the network architecture and communication capabilities of large-scale AI clusters require high-speed optical modules to be upgraded at an accelerated pace. According to Coherent's forecast, AI-related 800G and 1.6T digital optical modules are expected to account for nearly 60% of the total digital optical module market in the next few years, and the market size is expected to exceed 9 billion US dollars in 2028.

Changes in the computing power cluster network architecture have increased the flexibility of optical modules.

The network architecture of large-scale AI clusters needs to meet the requirements of large bandwidth, low latency, and lossless. The computing power cluster architecture changed from a leaf spine architecture to a fat tree architecture, and the unconvergent characteristics further amplify the demand for high-speed optical modules. Under PCIe5.0, H 100:800 G optical module = 1:2.5, and GH 200:800 G optical module = 1:9 under NVLink 4.0.

The optical module iteration cycle is accelerated, and attention is paid to the application of new materials and new technologies.

AI requires an accelerated iteration cycle of optical modules. From the 4th generation to the 2nd generation, the 1.6T optical module is expected to be gradually applied in 2024H2. As the single-channel rate of optical modules increases, power consumption, cost, and loss of electrical connection signals are also gradually increasing, and optical I/O interfaces are expected to further eliminate bandwidth bottlenecks for chip connections. Considering comprehensive cost reduction and efficiency, new packaging technologies such as LPO and CPO, as well as new materials such as silicon light and thin-film lithium niobate, are expected to accelerate application.

Competitive landscape: Domestic manufacturers lead, leading manufacturers benefit from AI at the core.

In the global optical module market, domestic manufacturers are already in a leading position. According to LightCounting data, among the top ten optical module manufacturers in 2022, the revenue of Chinese manufacturers has surpassed Western competitors. In the AI era, due to the accelerated iteration of optical modules, leading manufacturers with leading positions have advantages in R&D capabilities, scale advantages, supply chain capabilities, customer stickiness, and production capacity layout, and are expected to benefit the core. At the same time, changes in new technology are also having an impact on changes in the landscape, focusing on manufacturers with forward-looking layouts in fields such as LPO and silicon optics.

Risk warning: AI development and investment fall short of expectations; increased industry competition; global geopolitical risks; changes in the industrial chain caused by the development of new technologies.