After the bell Wednesday, Micron Technology (MU) posted fiscal Q3 revenue of $41.5 billion, more than quadrupling from $9.3 billion a year ago and crushing the Wall Street consensus of $35.8 billion.
Adjusted earnings per share came in at $25.11, beating the $20.60 estimate by 22%. Gross margins hit 84.9%, well above the already-record guidance of approximately 81%. The stock surged more than 9% in after-hours trading.
The headline numbers are staggering on their own. But the more important story for investors tracking the robotics theme is what CEO Sanjay Mehrotra said on the call - and what the company's automotive and embedded numbers quietly confirmed.
The Robotics Quote That Matters
Mehrotra has been building toward this moment for several quarters. Back in March, during Micron's Q2 call, he made a statement that deserves more attention than it received at the time:
Micron sees a "20-year growth vector in robotics" and expects the category to become "one of the largest product categories in the technology world."
Wednesday's call extended that thesis further - according to commentary on the post-earnings analyst call, Mehrotra noted that humanoid robots carry roughly 10x more memory than the average Level 2+ vehicle (L2+), setting up what Micron expects to be a substantial multi-decade demand cycle later this decade.
The company also expects L2+ and above vehicles to exceed 40% of the vehicle mix by 2030 - a near-term ramp that itself represents a significant memory demand event before humanoid robots even reach scale.
(A L2+ vehicle can handle steering, acceleration, and braking simultaneously while the driver remains engaged but with meaningfully reduced workload - think highway autopilot systems like Tesla's Autopilot or GM's Super Cruise.)
Wednesday's call extended that thesis. The capital expenditure expansion Mehrotra outlined - new fabs in Idaho, New York, Virginia, Taiwan, Singapore, and Japan - was framed not just around data center demand, but around the broader categories set to absorb memory over the next decade.
Analysts covering the call noted that management "insinuated these investments don't just support data center buildout, but other growing categories like self-driving cars, AI chips in smartphones and laptops, and potentially even robotics."
The robotics signal isn't hypothetical anymore. It's being built into the physical infrastructure.
The Automotive and Embedded Unit: The Leading Indicator
The clearest near-term data point for robotics investors isn't a speculative comment - it's a business unit line item. Micron's Automotive and Embedded Business Unit posted $4.63 billion in Q3 revenue, more than quadrupling year-over-year. Gross margins for the segment came in at 68%, up sharply from 45% in the prior quarter and just 21% two years ago.
That's the memory ecosystem that robotics will plug into. The same LPDDR5X and UFS 4.1 NAND products Micron is already shipping for advanced driver assistance systems are the building blocks for the memory stack in a humanoid robot.
The Level 4 autonomous vehicle - which Micron says requires over 300 gigabytes of DRAM compared to roughly 16 gigabytes in a standard car today - is functionally a peer of the compute platform a humanoid robot will require.
As that embedded segment scales, the pricing infrastructure, supply agreements, and chip architectures being built today are setting the table for robotics.
What Analysts Are Watching
Heading into this report, a key question for investors: whether robotics - "flagged as a 20-year demand vector" - is already being embedded in Micron's multi-year strategic customer agreements.
Micron now has 16 such agreements locked in for three to five years, covering an expected $100 billion in revenue backlog at floor pricing.
The implication: the memory supply chain for the categories adjacent to robotics is being locked up now, years before humanoid robot production scales. Companies that wait for robotics revenue to show up in an obvious line item will likely have missed the setup entirely.
The Bigger Picture
Micron guided Q4 revenue to $50 billion - roughly 16% above the $43.2 billion consensus - while signaling that supply shortages will persist "even as we expect industry supply to improve gradually in 2028."
The DRAM market is growing faster than previously anticipated, with management now guiding for low-to-mid 20% industry bit demand growth.
Every incremental robot that ships at scale is a customer for that DRAM. The clock on that demand is running.
And if you know anything about our philosophy here at NXT Wave Research, we want to position our portfolios into megatrends before Wall Street. The Robotics story just got a huge boost after the bell on Wednesday.
To your future success,
Matt McCall
Founder, NXT Wave Research
