Recently, the AI sector in the market has exploded, with two forces simultaneously gaining momentum. On one side, policy directives have clarified that artificial intelligence will be fully integrated into the education system, which means that AI+Education, AI+Healthcare, AI+Media and other niche tracks are all on the rise, backed by policy support.

On the other side, the industry is even more frantic—Meta has invested billions of dollars to acquire the domestic AI startup Manus, which is enough to show that the entire industry has entered a true competition phase. The giants are desperately pouring money to secure their positions.

There is a very important detail that many people haven't noticed: it seems that application development is booming, but the real money often flows upstream to hardware. CPO, PCB, liquid cooling equipment, optical chips—these directly support AI computing power—are the true targets of institutional funding. Why? Because the demand for these items is rigid and essential. If AI continues to explode in 2026, the certainty in hardware will far surpass the opportunities at the application level. This is a key investment logic to remember.