🔥 If we condense the tech megatrends of the next 5-10 years into 10 tracks, I believe the ones truly worth paying attention to are not the companies chasing hot topics, but the leading enterprises that control key links in the industrial chain, possess technological barriers, and have pricing power.

This is because every technological revolution ultimately follows the same path of development:

Surge in computing power demand → Growth in chip demand → Expansion of manufacturing equipment → Upgrade of advanced packaging → Expansion of optical communication → Follow-up of new materials.

And the following 10 companies almost cover the most critical positions of the entire chain.

1. InnoLight: The "Highway Toll Booth" of the AI Era

As AI training cluster sizes grow larger, data transmission between GPUs is becoming a new bottleneck.

From 800G to 1.6T, high-speed optical modules have become a necessity for AI data centers.

As one of the global leaders in optical modules, InnoLight is not only tied to top global cloud providers and AI computing power clients but also possesses industry-leading mass production capabilities.

While the market is still discussing computing power, InnoLight is already collecting fees for "connecting computing power."

2. YOFC: The Core Supplier of Optical Chips Behind Optical Modules

Without optical chips, there are no optical modules.

The greatest value of YOFC lies in its IDM model, achieving integrated control from design to manufacturing.

Against the backdrop of long-term dominance by overseas companies in optical chips, the space for domestic substitution remains vast.

Every additional AI server in the future means more demand for optical chips.

3. Cambricon: The Flag-Bearing Enterprise for Domestic AI Chips

The core of AI competition ultimately boils down to computing power competition.

Cambricon's greatest advantage is not short-term profits, but its strategic position within the domestic AI chip ecosystem.

As more and more enterprises seek domestic computing power solutions, Cambricon is becoming an important component of domestic AI infrastructure.

For the entire domestic computing power industrial chain, it is not just a company, but an ecosystem entry point.

4. Moore Threads: The New Force in Domestic GPUs

The market over the past few years has proven that one of the most important assets in the AI era is the GPU.

But GPU competition has never been just about chip competition.

Driver software, development tools, compatibility ecosystem, and application adaptation capabilities are equally important.

The core reason Moore Threads can continuously attract market attention is that it is building a complete GPU ecosystem, not a single hardware product.

5. JCET: A Beneficiary of the Advanced Packaging Era

In the past, chip performance improvements mainly relied on process advancements.

In the future, it will rely more on advanced packaging.

2.5D packaging, 3D packaging, and Chiplet architecture have become important development directions for high-end AI chips.

As more and more AI chips adopt advanced packaging solutions, the production capacity and technological value controlled by JCET will also increase accordingly.

Many people focus on chip design but overlook that packaging is becoming a new strategic high ground.

6. AMEC: The Leader in Domestic Semiconductor Equipment

Chips can be designed, but equipment is the hardest link in the industrial chain to break through.

AMEC's etching equipment has become an important representative in the field of domestic equipment.

More importantly, the company is transforming into a platform-type equipment enterprise, continuously expanding its product matrix.

Behind every new wafer fab, equipment investment is indispensable.

And equipment companies often have longer-term growth potential.

7. Uni-Trend Technology: An Underestimated Leader in Industrial Inspection

Any advanced manufacturing is inseparable from inspection.

Whether it's semiconductors, lithium batteries, or the new energy industry, quality control is a core link.

After achieving domestic production of its core X-ray light source, Uni-Trend Technology is opening up a larger market space.

This type of enterprise is often not in the spotlight, but it is an indispensable part of industrial upgrading.

8. Cathay Biotech: A Platform-Type Enterprise in New Materials

Technological competition is not just about chip competition.

Material innovation also determines the upper limit of an industry.

Cathay Biotech's greatest advantage lies in its ability to deploy across the entire industrial chain.

From strain R&D to final product, it forms a closed loop, possessing strong platform attributes.

In the long run, the new materials track often gives birth to companies capable of sustained growth.

9. Evergreen Technology: A Hidden Champion in High-End Chemical Materials

The development of many high-tech industries ultimately requires support from high-end materials.

The special monomer track where Evergreen Technology operates has high technical barriers and a relatively stable competitive landscape.

This type of enterprise may not become the hottest star stock in the market, but it often possesses more stable profitability and a longer growth cycle.

10. Hygon Information Technology: A Dual Beneficiary of Domestic CPUs and AI Computing Power

If GPUs determine AI training capabilities, then CPUs determine the foundational architecture of the entire computing platform.

Hygon Information Technology deploys both CPU and DCU product lines, holding a unique position within the domestic computing power system.

More importantly, software ecosystem and compatibility capabilities constitute extremely high barriers.

Chips can be copied, but ecosystems are hard to replicate.

This is also an important source of Hygon's long-term value.

Looking at the entire industrial chain, these 10 companies respectively occupy:

Optical Modules
Optical Chips
AI Chips
GPUs
Advanced Packaging
Semiconductor Equipment
Industrial Inspection
Biomaterials
High-End Chemicals
CPU Computing Power

Together, they form the most important underlying framework for China's tech industry upgrade.

Many investors are always looking for the next hot stock.

But truly great opportunities often come from anticipating industry trends early.

When AI, semiconductors, and advanced manufacturing enter a new round of capital expenditure cycles, the first to benefit are often not the end-user applications, but these enterprises that master core technologies and key links.

In the next 5 years, whoever can become the "infrastructure provider" in China's tech industry chain is more likely to achieve sustained value re-rating.