Wall Street Has a New Favorite AI Stock, and It Isn’t a Chip Designer. It’s the Company That Makes the Memory. They’re Calling Micron “the Next Nvidia.” Here’s the Bull Case — and the Bear Case Nobody Wants to Hear.
On June 28, 2026, TechCrunch ran a headline that would have sounded absurd two years ago: “Why Wall Street thinks US memory maker Micron is the next Nvidia.” Micron Technology (NASDAQ: MU) makes DRAM and NAND — the commodity memory chips that have, for three decades, been the most reliably boom-and-bust business in all of technology. Memory is the part of the computer nobody used to write love letters about.
Then the AI data-center buildout turned memory into the bottleneck. Every Nvidia GPU needs High Bandwidth Memory (HBM) bolted directly to it, and only three companies on Earth make HBM at scale. On June 24, 2026, Micron reported a fiscal third quarter that broke its own records on nearly every line — revenue more than quadrupled year over year — and guided the next quarter higher still. The stock hit an all-time high of $1,255 the next day.
This piece explains what the “next Nvidia” thesis actually rests on, walks through the numbers analysts are pointing to, and lays out the bear case those analysts mostly aren’t. It is explanatory business journalism, not investment advice. Nothing here is a recommendation to buy or sell any security.
- $41.46BFiscal Q3 2026 revenue — up 346% YoYMicron reported $41.46 billion in revenue for fiscal Q3 2026 (reported June 24, 2026) — up roughly 346% from $9.30 billion a year earlier and up 74% sequentially from $23.86 billion in Q2. The figure beat the Wall Street consensus of roughly $35–36 billion. Source: Micron 8-K / earnings release.
- ~85%Record gross marginNon-GAAP gross margin reached roughly 85% (GAAP 84.9%), up from about 39% a year earlier — a margin level that, for the quarter, ran ahead of even Nvidia's. Micron guided Q4 gross margin toward ~86%. Source: Micron prepared remarks; CNBC.
- $50BQ4 2026 revenue guidance — a recordManagement guided fiscal Q4 revenue to a record $50.0 billion (±$1.0 billion) with non-GAAP EPS of about $31.00 (±$1.00) — roughly $6–7 billion above analyst estimates going in. Source: Micron earnings call.
- Sold out2026 HBM capacity — fully committedMicron's entire calendar-2026 HBM output is pre-sold under multi-year contracts. The company disclosed 16 Strategic Customer Agreements; 14 of them carry cumulative minimum revenue commitments of roughly $100 billion over their three-to-five-year terms, backed by about $22 billion in customer cash deposits. CEO Sanjay Mehrotra said Micron can currently fill only half to two-thirds of HBM demand. Source: Micron; CNBC.
- ~325%MU stock — 2026 year-to-date gainMicron shares were up roughly 325% year-to-date and about 700% over the trailing 12 months as of late June 2026, hitting an all-time high near $1,255 on June 25 and pushing the company's market cap past $1 trillion. Past performance does not predict future results. Source: CNBC; Investing.com.
When analysts call Micron “the next Nvidia,” they are not saying Micron designs GPUs. They are making a narrower, structural argument: that the AI buildout has turned memory from a swappable commodity into a scarce, mission-critical input — the same role GPUs played when Nvidia’s revenue went vertical. An AI accelerator is useless without enough fast memory feeding it data. As models and context windows grow, the memory-to-compute ratio climbs, and the industry runs short of exactly the product Micron makes.
The clincher for the bulls is concentration. Only three firms produce HBM at scale — SK Hynix, Samsung, and Micron. That is an oligopoly, not a commodity free-for-all, and oligopolies hold pricing power when demand outruns supply. Micron’s data-center revenue exceeded $25 billion in the quarter — an annualized run rate north of $100 billion — from a product line that barely registered a few years ago.
Micron’s fiscal Q3 2026, reported June 24, was the quarter that converted skeptics. Revenue of $41.46 billion more than quadrupled from a year earlier and beat consensus by roughly $6 billion. Non-GAAP earnings of $25.11 per share blew past the ~$20.28 estimate. The driver was pricing: with HBM capacity scarce, DRAM average selling prices rose by a mid-60% range sequentially and NAND by a high-70% range, lifting gross margin to a company-record ~85%.
Then the guidance went higher. Management projected a record $50 billion in Q4 revenue — about $6–7 billion above what analysts had penciled in — with gross margin guided to ~86% and EPS near $31.00. The stock jumped about 15% the next day to a record. As CNBC put it, the memory crisis had pushed Micron’s margins past Nvidia’s and Meta’s, making the old commodity-memory maker, for one quarter at least, “tech’s new margin king.”
Memory is now a structurally scarce AI input, made by only three companies, sold out for a year, and locked under ~$100 billion of multi-year contracts with prepaid cash. Micron is a U.S.-based supplier with record margins and demand it physically cannot fully meet. If the AI buildout keeps its pace, the bulls argue, Micron’s earnings power has been permanently re-rated — the way Nvidia’s was — rather than temporarily inflated by a cycle.
High Bandwidth Memory stacks DRAM dies vertically and places them in the same package as the GPU, moving data far faster than conventional memory. It is also harder to make, consumes more wafer area per bit, and ties up DRAM capacity that would otherwise go to PCs and phones — which is why the AI memory crunch has begun raising prices on ordinary consumer devices, a phenomenon the trade press has nicknamed “RAMageddon.” Micron began shipping HBM4 for Nvidia’s next-generation “Vera Rubin” platform in March 2026 and says it is in high-volume production for its lead customer.
The scarcity is the whole story. When a supplier can fill only half to two-thirds of orders and customers are wiring billions in deposits to lock in future supply, the supplier sets the price. That is the dynamic analysts are extrapolating forward — and the dynamic the bear case says always, eventually, reverses.
We delivered record fiscal Q3 revenue of $41.46B with record gross margins, driven by unprecedented demand for our HBM and data center memory. Our 2026 HBM supply is sold out, and we are working to bring on capacity as fast as we responsibly can.
Micron soars on a blowout quarter as the AI boom fuels memory-chip demand — revenue quadruples year over year and the company guides next quarter to a record $50 billion.
Why Wall Street thinks US memory maker Micron is the next Nvidia: the AI data-center buildout turned memory into the bottleneck, and only three companies make HBM at scale.
Truth Social omitted — this is a non-political business story with no relevant official Truth Social presence; the social layer here is verified company and newsroom posting on X.
After the quarter, sell-side price targets were raised across the board, in several cases more than doubling. These are analyst opinions about a 12-month horizon, not facts and not guarantees — they move with sentiment and can be cut as fast as they are raised. With MU trading near $1,255 at its late-June high, the published targets clustered well above it.
Susquehanna: raised to $1,750 (from $600)
UBS: $1,625 · Aletheia Capital: $1,600
Needham: $1,550 (from $500) · Stifel / TD Cowen: $1,500
Bernstein / Wedbush: $1,300 · Wolfe Research: $1,250 · Citi: $1,200
Raymond James: $1,100 (from $530)
Consensus rating: Strong Buy; published median target near $1,087. Some bull-case scenarios reach above $1,500 on the assumption HBM demand extends through 2027. Source: aggregated sell-side notes via Investing.com, TheStreet, MarketWise.
Here is the part the price-target race tends to skip. Memory is the most reliably cyclical business in technology, and the cycle has the same shape every time: prices boom, makers over-invest in new fabs, supply floods in, prices crash, everyone bleeds, capacity gets cut, and shortage returns. Micron itself posted heavy losses in the 2023 downturn, barely three years before this record quarter. The bull thesis is, at bottom, a bet that “this time is different” — that AI demand is structural rather than another peak.
The specific risks: capacity. SK Hynix, Samsung, and Micron are all racing to add HBM and DRAM output, and a wave of new supply landing in late 2026 or 2027 just as order growth slows would flip today’s shortage into a glut and send pricing — and margins — back down hard. There is also demand pull-forward (hyperscalers stockpiling ahead of need), capex intensity (Micron is spending tens of billions on fabs that become stranded cost if utilization drops), and customer concentration in a handful of AI buyers. Even bulls who like the stock concede the cycle has not been repealed — RBC’s base case has the upcycle running through 2027, which is itself a bet on when, not whether, it turns.
“Every memory cycle ends the same way. The only question is whether AI has changed the demand curve enough to push the ending out by years — or whether investors are paying peak prices for peak earnings right before the peak.”
Synthesis of the bear case — per Investing.com, Blocks & Files, and Uncover Alpha analyses, May–June 2026
Civic Intelligence does not make stock recommendations. This article explains why a debate is happening, not which side is right. Stock prices, analyst targets, and multi-year revenue projections are forward-looking and speculative; memory pricing is historically volatile; and past performance — including Micron’s ~700% one-year run — does not predict future results. Anyone making financial decisions should consult a licensed professional and the company’s primary filings.
The “next Nvidia” label is a claim about position, not certainty. Micron sits on a genuine choke point in the AI supply chain, with record financials, an oligopoly’s pricing power, a sold-out year, and ~$100 billion of contracted demand behind it. That is the strongest version of the bull case the company has ever had. It is also a memory company, and memory has humbled every investor who assumed the good times were permanent. Both things are true at once — which is exactly why this is a story worth understanding rather than a stock tip worth chasing.
This is explanatory business journalism, not investment advice. Stock prices, analyst price targets, and revenue projections are forward-looking, speculative, and attributed to their sources throughout. Figures reflect public filings and reporting as of June 29, 2026 and will change.