Recently, there has been a phenomenon in US stocks that I think many people haven't realized yet.

AI is not over.

It's just that capital has started to shift from "only buying NVIDIA" to gradually spreading across the entire industry chain.

The recent performance of the storage sector is the best example.

Stocks like Micron (MU), Western Digital (WDC), SanDisk (SNDK), and Seagate (STX) have clearly become active during this period.

Many people might think that they are just rising along with AI.

I, on the contrary, believe that the market is starting to trade the second phase of AI.

In the first phase, everyone scrambled for GPUs.

In the second phase, they start scrambling for storage.

Why?

Because AI models are getting larger and inference is becoming more frequent. The real bottleneck to computing power is no longer just GPUs, but high-performance storage like HBM, DRAM, and NAND.

GPU is responsible for computation.

Storage is responsible for feeding data.

In the future, for every AI data center expansion, in addition to purchasing GPUs, a large amount of storage must also be purchased.

So now capital is starting to revalue the entire storage industry.

This is also why, after Micron's earnings report, even though short-term volatility is high, the market is still willing to trade it repeatedly.

Because what everyone is looking at is no longer how much money they made this quarter.

But whether AI capital expenditure can continue to grow in the next two to three years.

So, can you still buy now?

My answer is:

Yes, but not by blindly chasing.

This round of storage stocks has already accumulated a lot of gains. Chasing in the short term will definitely not have as good a risk-reward ratio as in the previous months.

If you are a short-term trader, I would rather wait for a pullback.

Because the semiconductor sector has a characteristic.

When it's rising, everyone is bullish; the truly comfortable buying point often appears when people start to doubt.

If you are long-term capital.

My view is actually simple.

As long as AI capital expenditure is not over, the boom cycle of the storage industry is not over.

There will definitely be adjustments in between.

But adjustments do not mean the logic is over.

The market's true wealth effect never comes from those stocks that hit new highs every day.

But from those stocks whose long-term logic hasn't changed but are mistakenly sold off due to sentiment.

If I had to rank the focus order for this sector.

First, I still look at Micron (MU).

Because it represents the most core high-bandwidth storage logic of the AI era, and it is also the only storage manufacturer in the US, with the highest certainty.

Second, I would look at Western Digital and SanDisk.

They are more like elastic plays. If enterprise SSD and NAND prices continue to recover, profit release speed may be faster than market expectations.

Third, Seagate.

It does not have as high growth potential as Micron, but with AI data center expansion, cold data storage demand will also continue to grow. It is more of a stable allocation.

Finally, let me state my view.

Now the market is no longer a question of "whether there is an AI rally."

But rather which sub-sectors the AI rally can still spread to.

Last year, the most profitable were GPUs.

Starting this year, more and more capital has turned its attention to storage.

If AI investment continues to increase in the future, I think the storage sector is still worth long-term tracking.

But I won't rush in because of a single green candle.

The real wealth effect is not achieved by chasing, but by daring to hold quality assets during pullbacks after the trend is confirmed.

$SNDK ‌$MU ‌#SK海力士ADR获超额认购