$SNDK is up 3,710% in 15 months. Most people still don't know what SanDisk actually does. Let me fix that.

The history, the explosion, the structural thesis — and why I'm waiting for the sea to break before I go all in. Everything you need to understand one of the most extraordinary stock stories of the decade.

What is SanDisk — actually?

You know SanDisk as the brand on the memory card in your camera or the USB stick in your drawer. That's the consumer face. The real business is something far more consequential: SanDisk is one of the world's largest manufacturers of NAND flash memory — the technology that stores data permanently in everything from your iPhone to the most powerful AI data centers on earth.

No moving parts. No power needed to retain data. Extraordinary speeds. As AI models grow larger and inference workloads multiply across every industry, NAND flash has gone from a commodity to a strategic asset. SanDisk makes it. And right now, the world cannot get enough of it.

The history: from USB sticks to the backbone of AI

SanDisk was founded in 1988 and pioneered consumer flash memory storage. In 2016, Western Digital acquired it for $19 billion — absorbing it entirely, burying the brand inside a conglomerate alongside hard disk drives, a technology slowly dying in the age of SSDs.

For nine years, SanDisk existed only as a division. Its potential hidden. Its upside diluted. Then, in October 2023, Western Digital announced the inevitable: it would spin off its flash business into a separate, independent company. On February 21, 2025, SanDisk Corporation began trading on Nasdaq under $SNDK — a pure-play flash memory company, free from the dead weight of spinning rust, singular focus on the technology the AI era actually needs.

It listed near $32. Today it trades near $1,468. That is not a typo.

Why it exploded: three things happened simultaneously

1. The spinoff unlocked the valuation. Buried inside Western Digital, SanDisk was priced like a commoditized hardware division. The moment it became a pure-play, the market re-rated it as a semiconductor and AI infrastructure company. Same business. Radically different multiple. This alone would have moved the stock significantly.

2. NAND prices went supernova. NAND flash prices surged 60% in Q1 2026 alone. Gartner projects a 234% annual price increase for the full year. No meaningful supply relief until late 2027 at the earliest. When you're one of three companies on earth that can produce NAND at scale, price inflation doesn't just improve your margins — it transforms your entire income statement overnight.

3. AI created a demand category that has never existed before. Every AI query, every model inference, every retrieval-augmented generation database, every vector store, every agent memory system — it all needs fast, high-capacity storage close to the compute. SanDisk ended 2025 as the top performer in the entire S&P 500 with a 559% gain. Then added another 494% in 2026 year-to-date. The stock has beaten Bitcoin over a 12-month period. By a factor of ten.

The structural case — even if AI disappoints

Here's what I find most compelling: you don't need AI to deliver everything the bulls are promising for this thesis to work. Global data center capacity is projected to nearly triple by 2030. Non-AI workloads alone — streaming, cloud storage, enterprise databases, 5G, connected devices — are growing at 11% CAGR. None of that needs a ChatGPT success story. All of it needs NAND flash.

And supply isn't catching up. NAND capex is growing at 5% this year while demand grows at multiples of that. Samsung and SK Hynix have both said publicly that shortages persist through 2028. Gartner says no pricing relief until late 2027. The gap between supply and demand is not closing — it's widening. SanDisk sits directly in the middle of that gap, with long-term supply contracts that include variable pricing clauses — meaning they capture future price increases automatically.

The one-sentence thesis

Everyone is talking about the chips that think. I'm investing in the chips that remember. There's more of the second kind needed — and the world is running out of supply through at least 2028.

⚠ Disclosure

I do not currently hold a position in $SNDK — and I'll be honest about why: I didn't pull the trigger when I should have. I had the thesis. I believed the numbers. I just didn't size the position I wanted to put on. That's on me, and I'm not going to pretend otherwise. I'm watching the stock carefully and I'm waiting for the sea to break — a meaningful pullback that gives me the entry I want. When that happens, I'm going in. The thesis hasn't changed. Only the price has.

Sources: Gartner Semiconductor Forecast (April 2026) · IDC Semiconductor Market Forecast (April 2026) · McKinsey Enterprise SSD Report · McKinsey Data Center Demand Model · IndexBox · StockAnalysis · Yahoo Finance. Not investment advice.