To The Moon

$MU $闪迪(SNDK)$ $美光科技(MU)$ $纳指100ETF(QQQ)$

After last night's earnings report, the biggest story wasn't the numbers.

It was what happened afterward.

Wall Street is starting to change the way it thinks about memory.

The most aggressive call came from DA Davidson.

They raised their Micron price target from $150 to $200 and maintained their Buy rating.

But the target itself isn't the interesting part.

The reason behind it is.

Their thesis can be summarized in three words:

New Era in Memory.

And that may be the most important takeaway from this entire earnings cycle.

For decades, memory was viewed as one of the most cyclical industries in technology.

Prices rise.

Profits surge.

Supply expands.

Prices collapse.

Repeat.

Investors learned to treat memory stocks like commodities.

Micron is now arguing that the AI era is changing that model.

And the latest earnings report provides some compelling evidence.

First, Micron didn't just beat expectations.

It completely shattered them.

Q4 revenue guidance came in at $4.9B-$5.1B versus Wall Street expectations of roughly $4.29B.

EPS guidance landed at $3.00-$3.20 versus expectations around $2.55.

That's not a beat.

That's a complete reset of expectations.

More importantly, management made something very clear:

They still do not see memory supply catching up with demand anytime soon.

Think about that.

For years, every memory cycle ended the same way.

Supply eventually overwhelmed demand.

This time, the story looks very different.

HBM is consuming massive amounts of advanced DRAM capacity.

Leading-edge manufacturing is becoming increasingly complex.

New fabs require enormous capital and years to build.

Meanwhile demand continues accelerating.

AI training.

AI inference.

Agentic AI.

Long-context models.

Enterprise AI deployments.

Every major trend in artificial intelligence is driving memory consumption higher.

The market continues asking when memory pricing will peak.

Micron seems to be asking a different question:

What if this isn't a normal cycle?

Then there's the most overlooked announcement from the earnings call.

Micron has now signed 16 Strategic Customer Agreements.

These agreements span data centers, consumer electronics, and automotive markets.

Some extend all the way to 2030.

Many include take-or-pay provisions.

In simple terms:

Customers are reserving future capacity years in advance.

And in some cases, they pay regardless of utilization.

That is not how commodity industries behave.

That is how strategic resources behave.

The implication is enormous.

Customers are no longer worried about paying too much.

They're worried about not getting enough supply.

And that's a very different market structure.

The demand story is also expanding far beyond AI servers.

Autonomous vehicles require significantly more memory than traditional vehicles.

AI PCs require more memory.

AI smartphones require more memory.

Edge AI devices require more memory.

Humanoid robots could eventually require multiples of the memory content found in advanced vehicles.

For decades, memory demand was driven primarily by PCs and smartphones.

The next decade could be driven by AI infrastructure, autonomous systems, robotics, and entirely new computing platforms.

That's not a cyclical story.

That's a platform expansion story.

Which brings us back to the DA Davidson upgrade.

The reason they raised their target isn't because Micron had a good quarter.

The reason is that they believe earnings visibility is reaching levels the memory industry has never seen before.

And if they're right, Wall Street may have to rethink how memory companies are valued.

Because if memory is no longer just another commodity business...

Then Micron is no longer just another memory stock.

For years, investors believed NVIDIA was the most important asset in AI.

NVIDIA still drives demand.

But increasingly, memory is becoming the mechanism that validates demand.

Because when GPUs are sold out, it's a headline.

When memory is sold out, it usually means the entire AI infrastructure ecosystem is expanding.

That's why my framework remains unchanged:

Memory first.

Optical interconnect second.

Applications third.

And after this earnings report, I see even more evidence supporting that view.

The AI boom isn't running out of memory.

The AI boom may be turning memory into one of the most valuable assets in the entire technology stack.