Wall Street's Boldest Price Target Is $1,625 for Micron. Here's Why the $1 Trillion Milestone Might Be Just the Opening Act.Wall Street's Boldest Price Target Is $1,625 for Micron. Here's Why the $1 Trillion Milestone Might Be Just the Opening Act. The last time a memory chip company joined the $1 trillion club, most analysts didn't see it coming. Micron Technology (MU) crossed that line Tuesday, but the market cap figure isn't the real story. The real story is UBS. The firm just raised its price target for Micron from $535 to $1,625. That's the highest price target on Wall Street right now, and it implies another 71% upside from where the stock trades today near $950. The upgrade sent shares up nearly 18% in a single session, pushing the market cap to $1.07 trillion. If you think that sounds aggressive, you're not alone. But here's what UBS is looking at. ## The AI Memory Crunch Nobody Talks About Most of the AI infrastructure conversation centers on Nvidia. The chips. The GPUs. The data centers. What doesn't get as much attention is the memory those systems require. Every Nvidia H100, H200, and Blackwell accelerator runs on high-bandwidth memory, or HBM. It's a specialized type of DRAM that stacks vertically and connects directly to the processor, delivering data at speeds that conventional memory can't match. Without it, the AI chip doesn't perform anywhere near its spec sheet. There are exactly three companies in the world with the capability to produce it at scale. SK Hynix. Samsung. And Micron. Micron has already sold out its entire 2026 supply of HBM. The year isn't halfway done, and every unit is spoken for. That kind of demand picture doesn't look like a cyclical memory cycle. It looks like a structural shift. "AI is transforming memory into a critical bottleneck," one analyst wrote this week. That framing matters. For decades, Micron was viewed as a commodity business, a company that made things when demand was high and bled when it wasn't. The entire investment thesis was built around timing the cycle. That model is being retired. ## Why the Business Looks Different This Time The shift shows up in the contract structure. Micron has been moving toward long-term agreements with partially fixed pricing for its AI-focused memory products. That's not what commodity chipmakers do. That's what infrastructure companies do. For investors, this changes the earnings model. Instead of trying to predict where memory prices go in the next two quarters, the question becomes how fast the AI buildout continues. And on that question, the answer from every major hyperscaler, every cloud provider, and every government AI initiative points in one direction. UBS projects Micron's valuation could approach $1.8 trillion within 12 months. That's roughly where Nvidia was before its most recent leg up. It requires memory to stay sold out, pricing to hold, and AI spending to continue accelerating. All three of those things are more likely than they were two years ago. ## The Risk Is Real None of this means the trade is risk-free. Micron has been one of Wall Street's most reliable boom-bust stories. The company's earnings historically swing from record profits to operating losses within a few quarters, depending on where memory prices are in the cycle. Investors who have owned this stock for a decade know what a down cycle looks like, and it's not pretty. The stock has already tripled in 2026. At $950 a share, it's priced for a lot of good news. A slowdown in AI infrastructure spending, a faster-than-expected production ramp from SK Hynix or Samsung, or any wobble in enterprise cloud budgets could send shares back toward the mid-$600s quickly. The other thing worth watching: UBS's $1,625 target is the most aggressive on the Street by a significant margin. Most analysts cover Micron with more conservative assumptions. When one firm is that far above consensus, it either means they see something the others don't, or they're wrong. ## Bottom Line Micron at $950 sits 71% below UBS's 12-month target. The company has achieved something its bears never expected: sold-out supply in a market that used to price memory like a commodity. The $1 trillion milestone is real. So is the risk of a 130-year-old cyclical thesis reasserting itself at the worst possible time. If the AI memory thesis holds, Micron's upside is significant. If it doesn't, the stock is priced for disappointment on the way down. Investors entering here are essentially making a bet on the duration of the AI infrastructure buildout. Right now, the market is voting yes.