Forget Nvidia and Taiwan Semiconductor. This $1 Trillion Memory Play Just Soared 90% in 2.5 Months.

Photo of Jeremy Phillips
By Jeremy Phillips Published
This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.

Everyone’s chasing obvious AI semiconductor names. NVIDIA (NASDAQ:NVDA | NVDA Price Prediction) sits at a $5.15 trillion market cap. Taiwan Semiconductor Manufacturing (NYSE:TSM) is up 39% year to date. And Micron Technology (NASDAQ:MU) just crossed $1 trillion in market cap, joining SK Hynix and Samsung in a memory club that suddenly looks unkillable.

On a recent episode of The AI Investor Podcast, Eric Bleeker and Austin Smith argued the better trade is the company selling picks and shovels to every memory fab on the planet, up 91% in just 2.5 months. Here is the case.

The $1 Trillion Memory Backdrop

Micron’s fiscal Q1 2026 report set records across the board. Revenue of $13.64 billion, up 57% year over year. The Cloud Memory unit nearly doubled to $5.28 billion with a 66% gross margin and 55% operating margin. Q2 guidance is $18.70 billion with a 68% non-GAAP gross margin.

Bleeker on the podcast: “Memory, Micron, SK Hynix, Samsung, they’re all over a trillion dollars right now because they’re getting 80% margins. There are companies that could go back down to 60% and these companies still win in a big way.” The supercycle is real. The question is how to position without paying for stocks that have already tripled.

ACM Research: 91% in 2.5 Months

The AI Investor Podcast recommended ACM Research (NASDAQ:ACMR) on March 13. The stock has climbed from $46.38 to $88.64 since, up 125% year to date.

ACM supplies cleaning, electroplating, furnace, and advanced packaging tools to YMTC and CXMT, the two Chinese memory makers building domestic capacity. Bleeker called the company “a great company that’s really rock solid and is positioned in an ideal place.” His view on the build cycle: “China can’t build enough memory for itself, let alone to flood the Western market.”

Q1 2026 revenue came in at $231 million, up 34% year over year, with total shipments of $241 million, up 54%. Management maintained full-year guidance of $1.08 billion to $1.18 billion, or 21% to 30% growth and reaffirmed a $4 billion long-term revenue target. The ECP segment nearly tripled to $84 million, driven by advanced packaging adoption.

The regulatory moat matters here. China lacks access to EUV lithography systems for sub-5nm production, so the AI build happens through the layers ACMR serves: cleaning, plating, 3D packaging. That is a multi-year capex tailwind independent of any single customer hitting targets.

Onto Innovation: The Advanced Packaging Angle

Onto Innovation (NYSE:ONTO) is the second leg of the trade, up 19% since late February. Onto makes metrology and inspection tools for advanced packaging, where innovation is moving as HBM stacks grow taller and 3D integration replaces traditional node shrinks. Q4 2025 set a record at $267 million, with Q1 2026 guidance of $275 to $285 million and an HBM-related volume agreement worth over $240 million through 2027.

The Old Tech Premium

Austin Smith raised a wrinkle worth flagging. In a chip-starved world, any chip could be put to use. He pointed to legacy CPU inventory as evidence that old tech can command a premium. GlobalFoundries (NASDAQ:GFS) is up 132%.

The surprise: United Microelectronics (NYSE:UMC), a Taiwanese fab focused on older nodes, is up 183% year to date, outpacing Taiwan Semi. Bleeker on the podcast: “In a perverse way, buying the stuff with the older facilities has outperformed Taiwan Semiconductor this year.” Many ways to win.

What I’m Watching

I have followed the semi cycle for over a decade and the lesson keeps repeating. When obvious names go vertical, second-derivative plays often have more runway. Bleeker’s read is that equipment makers like ACM Research and Onto Innovation have not run up like memory and GPU names. If Micron’s order book signals anything, the supercycle stretches into 2027, and that gap could close.

You should weigh ACMR if you believe Chinese memory capacity has years left to build. You should pass if tariff escalation or a memory glut arrives before YMTC and CXMT finish their fabs. Either way, the trillion-dollar memory club is no longer the whole story.

Photo of Jeremy Phillips
About the Author Jeremy Phillips →

I've been writing about stocks and personal finance for 20+ years. I believe all great companies are tech companies in the long run, and I invest accordingly.

Continue Reading

Top Gaining Stocks

AMAT Vol: 4,466,729
WDC Vol: 5,750,522
MRNA Vol: 2,345,977
SNPS Vol: 597,617
AVGO Vol: 14,912,447

Top Losing Stocks

CTRA Vol: 73,319,495
KMX Vol: 3,511,741
CME Vol: 2,093,954
HP
HPQ Vol: 6,005,099
IQV Vol: 637,168