When the artificial intelligence (AI) boom started gathering momentum in early 2023, Micron Technology (MU +0.04%) stock was trading at around $50. Last month, it peaked at $1,213, representing a staggering 2,070% gain over the past three and a half years.
Micron’s high-bandwidth memory (HBM) for the data center is a critical component in the AI hardware stack, and demand currently exceeds supply, which is driving explosive growth in the company’s revenue and earnings. However, recent reports suggest some businesses are curbing their AI software spending, which could affect future demand for chips and other hardware components.
As a result, Micron stock has plummeted by 30% from its recent high and closed at $848 on Friday, July 17. Should investors buy it while it’s trading under $1,000? Read on for the surprising answer.
Image source: Getty Images.
Micron’s HBM is crucial to the AI revolution
AI training and inference workloads require an astronomical amount of computing power, which is delivered by specialized data center chips called graphics processing units (GPUs). HBM stores information in a ready state for when GPUs are ready to process it, which keeps workloads flowing smoothly. Without sufficient memory capacity, GPUs would have to pause while they wait for more data, creating a sluggish experience for anyone using an AI chatbot or agent.
Micron recently started shipping its new HBM4 chips for the data center, which provide a 60% increase in capacity and a 20% improvement in energy efficiency over its previous HBM3E chips. In other words, HBM4 is designed to maximize processing speeds and minimize costs, which is why Nvidia has adopted this solution for its new Vera Rubin GPU systems.
The market for data center HBM was worth $35 billion last year, but Micron expects it to nearly triple to $100 billion by 2028, so this is a significant financial opportunity. But it isn’t Micron’s only opportunity in the AI space, because it also supplies memory and storage chips for personal computers, smartphones, cars, and even robots.
AI models are slowly becoming more efficient, so computers and smartphones can run them locally without relying on external computing power from data centers, as long as they have an appropriate amount of memory. Furthermore, the average car with even basic autonomous driving capabilities needs more than five times the memory capacity of a regular car — and Micron says humanoid robots need 10 times more memory than the average autonomous car.
Simply put, this could be one of the longest demand cycles for memory that Micron has ever experienced, so its stock looks like a screaming buy from that perspective. But there is room for caution, and I’ll explain why in a moment.
Today’s Change
(0.04%) $0.37
Current Price
$853.57
Key Data Points
Market Cap
Day’s Range
$804.01 – $903.88
52wk Range
$103.38 – $1255.00
Volume
9.5K
Avg Vol
51.8M
Gross Margin
72.60%
Dividend Yield
0.06%
Micron’s sales are growing at a mind-boggling rate
Micron delivered a record $41.4 billion in revenue during its fiscal 2026 third quarter (ended May 28), a whopping 346% increase from the year-ago period. AI-related memory fueled that strong result across all four of the company’s business segments, led by cloud memory, which contributed the most revenue thanks to booming HBM sales.
Micron’s earnings also rocketed higher by 1,368% year over year to $24.67 per share during the quarter. Management’s forecast for the current fourth quarter (which ends in late August) points to more record results, with $50 billion in revenue and $30.73 per share in earnings potentially in the cards.
The global memory shortage is giving Micron and its competitors the ability to dictate prices, which is significantly boosting the company’s revenue and profit margins. But every major memory company is frantically building more manufacturing capacity, so supply will eventually catch up to demand, which will make it very difficult for the likes of Micron to maintain its current level of earnings. As a result, its stock isn’t a clear-cut buy based on its recent financial results alone.
Should you buy Micron stock while it’s under $1,000?
As memory supply inevitably increases over the next couple of years, it might run into a simultaneous decrease in demand. Soaring infrastructure costs have forced AI providers such as Anthropic and Microsoft to implement passive price increases for the use of their models and software, so many of their customers are rethinking their usage.
A recent survey from investment bank UBS Group found that 60% of businesses are curbing their AI spending by routing some tasks to cheaper, more efficient models. Even the biggest companies are feeling the pinch; Walmart, Amazon, and Uber Technologies have recently capped AI usage for their employees to prevent budget blowouts.
Uber’s chief operating officer recently said it’s getting harder to justify AI spending, after his company burned through its entire 2026 budget in just four months by using Anthropic’s Claude Code. This doesn’t bode well for long-term semiconductor demand.
Micron stock looks like a bargain given its price-to-earnings (P/E) ratio is just 19.2, making it substantially cheaper than the Nasdaq-100 index which has a P/E of 33.4. Plus, based on Wall Street’s earnings estimate for fiscal 2027, Micron’s forward P/E is just 5.6. In my opinion, that suggests investors aren’t totally convinced the memory boom has legs, because a company growing this fast would normally command a premium valuation compared to the broader market, not a steep discount.
Micron stock could deliver a positive return from here over the next five years or so, once physical AI segments such as autonomous vehicles and robotics are commercialized at scale. But there could be some major volatility in the near term as additional memory supply comes online and the demand picture becomes more uncertain, so I’m in no hurry to buy Micron’s recent dip.
