Micron Is Now the Third-Most Profitable Company in America — How to See the Real AI Memory Bet in Your Sheet
Micron just printed $41.46 billion in quarterly revenue, an 84.9 percent gross margin, and $28.24 billion in net income. Only Nvidia and Google sit ahead of it in U.S. profit rankings. The HBM-driven memory cycle has turned a once-cyclical chipmaker into a member of a three-company AI infrastructure club, and here is how to size your real exposure across every brokerage you use in one Google Sheet.
The Numbers at a Glance
| Q3 FY2026 Revenue | $41.46B (vs $35.84B est) |
| Year-Ago Q3 Revenue | $9.3B — more than quadrupled |
| Q3 Adjusted EPS | $25.11 (vs $20.78 est) |
| Q3 Net Income | $28.24B / $24.46 per share GAAP |
| Gross Margin | 84.9% (vs 74.9% prior qtr, 39% year ago) |
| Q4 Revenue Guide | ~$50B (vs $43.58B est) |
| Locked-In Contracts | 16 long-term deals, ~$22B committed |
| Market Cap | Past $1T in May 2026; stock up ~700% in 12 mo |
| U.S. Profit Rank | Third, behind only Nvidia and Google |
Micron’s fiscal Q3 2026 numbers were not just a beat. They were a structural shift in what kind of company Micron is. An 84.9 percent gross margin is no longer a memory cycle peak. It is software-style economics, sustained by High Bandwidth Memory (HBM) sold into the build-out of Nvidia, Google, and hyperscaler AI clusters. The company’s market cap crossed $1 trillion in May, and the stock ran another 15 percent in extended trading on the print. By profitability, only Nvidia and Google now sit ahead of MU on the U.S. corporate leaderboard.
Why the Boom Is Not a One-Quarter Story
The part that should worry the bears is the visibility. Micron disclosed 16 long-term customer agreements — with data center operators, automakers, and AI platform builders — covering three to five years of committed volume. Roughly half of the company’s total revenue is expected to flow through these contracts once they are fully executed, with about $22 billion in financial commitments already signed. That is not a forecast. That is contracted demand. CFO Mark Murphy told analysts the agreements give Micron visibility to make capital investments with confidence, because the volumes are committed before the fabs are built.
CEO Sanjay Mehrotra added that memory and storage supply shortages “will take considerable time to improve, even as we expect industry supply to improve gradually in 2028.” Translated: the bottleneck is not pricing. It is wafer capacity. New fabs in Idaho and New York are years away from peak output, and the lead times on advanced HBM are already sold out through 2027. Memory has moved from a cyclical commodity business to a contracted infrastructure input, and the cycle math has changed with it.
The unit-economics story is the one the equity market is repricing fastest. A memory chip that used to compete on bits-per-dollar now competes on bandwidth-per-watt and proximity to the GPU. HBM3E and HBM4 are co-designed with Nvidia’s Rubin and Google’s TPU generations. The customers cannot easily swap suppliers without re-qualifying an entire accelerator platform, which gives Micron real pricing power for the first time in the company’s history.
Where the Growth Is Coming From
Every business unit at Micron grew sharply in the quarter, but the data center engine is doing the heavy lifting. Cloud memory revenue rose more than 300 percent year over year to $13.77 billion. Core data center sales climbed more than sevenfold to $11.5 billion from $1.53 billion a year ago, and the company also recorded more than $5 billion in data center SSD revenue. The mobile and client business grew 250 percent to $11.52 billion, and even the long-cycle automotive and embedded segment more than quadrupled to $4.63 billion. The story is not a single product line. It is memory in general becoming the bottleneck for AI deployment.
That breadth matters because it answers the most common bear case. A single-product boom (a new iPhone cycle, a Windows refresh, a datacenter buildout) is fragile. A cross-cycle shortage where every memory-consuming industry — phones, PCs, cars, servers, AI accelerators — is competing for the same fabs is a different animal. It is what DRAM bulls in 2017 and 2018 thought they were seeing, and it is why this cycle is being priced more like a platform transition than a commodity upturn.
Where Micron Sits in Your Real Portfolio
The single biggest gap between what investors think they own and what they actually own is in stocks that have moved this much, this fast. A 700 percent run over twelve months turns a small starter position into a top-three holding — except that most retail investors hold it in three places (their Roth, their taxable brokerage, and their spouse’s IRA), and none of the brokerage apps roll those up. The percentage of your net worth sitting in MU today is almost certainly higher than the percentage you targeted when you bought it. That gap is what the next bear market punishes first.
The fix is the same one that works for every concentrated AI trade: pull every brokerage into one Google Sheet, run your exposure formulas, and see the number. InvestSheet syncs Fidelity, Schwab, Robinhood, Vanguard, IRAs, and 401(k)s into one place and exposes built-in IVS_BROKERAGE formulas for the numbers that matter.
What to Watch Next
Three numbers will decide whether the boom extends into 2027 or peaks here. First, the Q4 revenue guide of $50 billion. The street had modeled $43.58 billion, so the bar is already high. A clean beat at that level confirms the contracted demand is converting into shipped product. Second, HBM mix as a percentage of DRAM revenue. Every percentage point of mix shift toward HBM compounds the gross margin, because HBM carries roughly three to four times the per-bit pricing of standard DDR5. Third, capacity announcements from competing suppliers — primarily SK Hynix and Samsung — and whether they signal relief in late 2027 or early 2028. Mehrotra has already telegraphed gradual improvement in 2028, so the question is not whether supply normalizes but how quickly.
None of that requires guessing the next twelve months. It requires watching the numbers as they print, knowing your real cost basis, and being honest about what percentage of your net worth is sitting in a single memory chipmaker. A position that quadruples in twelve months is a position that can halve in the next twelve — especially when the broader index is paying roughly 1 percent to wait. Make sure your sheet shows you the number before the next earnings call does.
Frequently Asked Questions
How profitable is Micron right now?
Micron reported fiscal Q3 2026 revenue of $41.46 billion against an LSEG consensus estimate of $35.84 billion, and adjusted EPS of $25.11 versus $20.78 expected. Net income came in at $28.24 billion, or $24.46 per share on a GAAP basis, up from $1.89 billion a year earlier. Gross margin reached 84.9 percent, compared with 74.9 percent in the prior quarter and 39 percent a year ago. By U.S. profit rankings, only Nvidia and Google now sit ahead of Micron.
Why is Micron suddenly so profitable?
High Bandwidth Memory (HBM) is the driver. HBM is a specialized DRAM architecture stacked directly onto AI accelerator chips such as Nvidia GPUs, and demand from data center buildouts has outrun supply. Micron has signed 16 long-term customer agreements worth roughly $22 billion in committed revenue, with terms of three to five years. About half of Micron’s total revenue is expected to flow through these contracts once they are fully executed, locking in visibility through 2028 even as supply gradually ramps.
How do I track my Micron position across multiple brokerages in Google Sheets?
Sync every brokerage that holds Micron — Fidelity, Schwab, Robinhood, Vanguard, IRA, 401(k) — into a single Google Sheet, then use IVS_BROKERAGE formulas to surface your real exposure. =IVS_BROKERAGE(“qty”, “MU”) returns your total shares across all linked accounts, =IVS_BROKERAGE(“value”, “MU”) returns the dollar value of the position, =IVS_BROKERAGE(“gainLoss”, “MU”) returns the unrealized gain or loss, and =IVS_BROKERAGE(“costBasis”, “MU”) returns the basis. From there, divide the MU value by =IVS_BROKERAGE(“value”) to see what percentage of your whole portfolio the position now represents — a number that has likely grown sharply as the stock has run up roughly 700 percent over the past year.
See your real Micron position across every account
14-day free trial. $9.99/mo or $7.99/mo annual. Sync 35+ brokerages — including Fidelity, Schwab, and Robinhood — into a single Google Sheet with live formulas for shares, value, cost basis, and unrealized gain on every ticker at once.
Install InvestSheet