Micron is no longer being treated as just another cyclical memory stock. UBS now sees AI demand and long-term supply deals pushing the company toward a valuation normally reserved for the largest technology names.
Micron Technology became the stock market's loudest AI signal on Tuesday, after UBS more than tripled its price target and argued that the memory maker could be worth close to $1.8 trillion within the next year. That is not a small adjustment. It is Wall Street saying the AI trade has moved beyond GPUs and into the parts of the system that make those GPUs useful.
According to a report from Reuters, UBS lifted its target on Micron to $1,625 from $535, the highest among 46 brokerages covering the company, after the stock closed Friday at $751. The call sent Micron shares up about 14% in early trading on May 26 and pushed its market value closer to $1 trillion from roughly $846.93 billion at the prior close.
The numbers are striking, but the reasoning matters more. UBS is betting that high-bandwidth memory, or HBM, has changed the earnings profile of a company investors used to punish for the old memory cycle. In the past, DRAM and NAND were classic boom-and-bust businesses. Demand rose, suppliers added capacity, pricing cracked, and investors waited for the next downturn. AI infrastructure is testing that pattern.
HBM is not ordinary memory. It is stacked, power-efficient DRAM built to sit close to accelerators from Nvidia and other AI chipmakers, feeding them data fast enough to train and run large models. Without it, an expensive AI processor is still expensive, but less useful. That is why memory has become one of the most important constraints in data centers.
Micron has a rare position in that market. It is the only major U.S.-based memory manufacturer competing in HBM against SK hynix and Samsung, and it has moved deeper into Nvidia's next-generation platform. Micron said in March that it had begun volume shipments of its 36GB 12-layer HBM4 product in the first quarter of calendar 2026, designed for Nvidia's Vera Rubin systems. That detail explains why investors are giving the stock a different look.
When the AI market was mostly discussed through Nvidia, the investment case was simple. Buy the company selling the accelerators. But a GPU cluster is not one product. It is a system of compute, memory, networking, storage, power, cooling and long-term supply planning. If any one of those pieces is tight, the whole buildout slows. HBM has become one of those pieces.
This is where the UBS call becomes more than a price target. The firm is arguing that long-term agreements across the memory industry could lock in volumes and partially fix pricing, giving Micron more visibility than investors usually expect from a memory supplier. Hyperscalers appear increasingly willing to give up some pricing flexibility in return for assured supply. That is what scarcity does. It changes the negotiation.
The valuation question has changed
Micron was trading at about 8.4 times expected earnings over the next 12 months, Reuters reported, compared with 21.1 times for the S&P 500 and 24.66 times for the Nasdaq 100. UBS's point is that this gap becomes harder to defend if Micron's earnings are no longer as volatile as the market assumes.
That is a bold argument, because memory investors have heard versions of this before. Every strong cycle produces claims that the industry has become more disciplined. Then supply catches up, prices fall, and discipline suddenly looks less permanent. The difference this time is that HBM is more specialized, harder to ramp quickly, and tied directly to multi-year AI capital spending plans from cloud companies that cannot afford to miss capacity windows.
Still, a $1.8 trillion Micron would carry a heavy burden of proof. It would place the company near the largest U.S. technology companies and imply that investors are willing to value memory more like core AI infrastructure than a commodity component. That may prove right, but it depends on HBM pricing staying firm, Nvidia and other accelerator vendors keeping demand high, and competitors failing to flood the market with supply.
The broader message for the AI trade is clear. The market is starting to reward companies that control scarce pieces of the infrastructure stack, even if those companies were not the first names investors bought when generative AI took off. Lam Research, Marvell, AMD, Qualcomm and other chip names also moved higher Tuesday, but Micron was the cleanest expression of the idea that the AI buildout is still constrained by physical supply.
For startup founders and enterprise buyers, there is a practical lesson here as well. AI costs are not just about model access or cloud credits. They are tied to a hardware chain where memory availability can shape server pricing, deployment timelines and the economics of inference. If HBM remains tight, the companies closest to supply will have leverage over everyone building on top of it.
Micron's rally may cool, because stocks that move this far this fast usually invite profit-taking. But the investment thesis has moved into a new phase. The next thing to watch is not simply whether UBS is right on $1,625. It is whether long-term HBM contracts keep turning memory from a cyclical trade into a strategic AI asset.
Also read: Qualcomm wins a ByteDance AI chip deal as scrutiny tightens • xAI is treating Cursor like a rival it cannot ignore • Stord raises $250 million as brands rethink Amazon fulfillment
