To The Moon

Everyone knows the AI infrastructure story:

More AI → more GPUs → more data centres.

But there is another part of the equation that is getting increasingly difficult to ignore:

Memory.

And I think the most interesting question now isn’t simply which memory stock can go higher?

It is:

How long can the memory shortage last before high prices create the supply that eventually ends the boom?

That is a very different question.

Recent trading has put memory stocks back in the spotlight. $SNDK jumped roughly 11% Friday, while other semiconductor names also rallied as investors focused on AI-driven memory demand. 

At the same time, Intel’s CEO has warned that memory prices have surged dramatically, highlighting just how tight the market has become. 

The bullish case is straightforward.

AI servers need enormous amounts of memory and storage.

Nvidia is now expecting to sell twice as many chips next year as this year, according to CEO Jensen Huang. While that is a forecast rather than confirmed orders, it reinforces the argument that AI infrastructure demand is still expanding rapidly. 

And memory suppliers are in a position that is very different from the early stages of the AI boom.

💰 The shortage creates pricing power.

If demand is rising faster than supply, manufacturers can raise prices without necessarily increasing production volumes at the same rate.

That can create a powerful earnings cycle for companies such as:

💾 $MU — Micron

💾 $SNDK — SanDisk

💾 $SK Hynix

💾 $WDC — Western Digital

But this is where the story gets interesting.

⚠️ High prices are also the cure for high prices.

When memory becomes extremely profitable, competitors have a much stronger incentive to invest in capacity.

And that is already starting to happen.

Reuters reported Friday that China’s CXMT is planning to expand into NAND flash memory, targeting the growing demand from AI servers. Meanwhile, SK Hynix’s Solidigm unit is considering a potential U.S. NAND manufacturing facility. Neither development means additional supply arrives immediately, but both show that the economics are attracting new investment. 

This creates a fascinating investment cycle:

AI demand ↑

↓

Memory demand ↑

↓

Prices ↑

↓

Memory profits ↑

↓

Capacity investment ↑

↓

Eventually, supply catches up

The timing is the difficult part.

If supply remains constrained for longer than expected, memory companies could continue benefiting from pricing power.

But if new capacity arrives while AI spending slows, the same stocks that benefited from scarcity could face a very different environment.

And there is another risk I think investors should watch:

Valuation can move faster than fundamentals.

A shortage can create spectacular earnings growth.

But markets often price the future before the financial statements catch up.

That’s why I don’t think the memory story should simply be viewed as:

“AI is booming, therefore memory stocks go up.”

The better question is whether AI demand can grow faster than memory supply for long enough to create a sustained earnings cycle.

That is the part I’m watching.

🔥 Is the current memory shortage the beginning of a multi-year super-cycle — or are high prices already sowing the seeds of the next oversupply?