Micron’s stock performance is drawing comparisons to Nvidia’s, but historical trends suggest caution. (Illustrative AI-generated image).
- Micron has secured $100 billion in future sales, primarily for high-bandwidth memory (HBM) crucial for AI systems.
- HBM is a specialized memory type essential for the speed required by AI training and large language models.
- Micron is one of only three global HBM suppliers, positioning it as a key US-based player in AI hardware.
- The memory chip industry is historically cyclical, with booms often leading to busts due to overbuilding and demand fluctuations.
- While AI demand is robust, a Harvard expert cautions that this cycle might not be different, citing past booms that eventually ended.
- Competition is increasing as Intel and AMD develop AI accelerators, potentially altering the market dynamics for memory suppliers like Micron.
Wall Street is whispering a new name alongside Nvidia: Micron. But history says to be careful.
The US memory chip maker has suddenly become one of the hottest stocks in the AI boom. Investors who missed Nvidia’s record-shattering run are now betting that Micron will be the next big winner. And on paper, the story is compelling. Micron has locked in $100 billion in future sales, mostly for the high-speed memory chips that AI systems need to function. Some analysts are even declaring that the old boom-bust cycle for memory chips is dead for good.
But a closer look at the numbers and the history of the memory industry suggests that investors should not get too comfortable. A Harvard chip expert warns that “this too will pass,” pointing to decades of evidence that memory demand goes in cycles. Meanwhile, a shift is happening in AI hardware. Intel and AMD are gaining ground while Nvidia’s growth is slowing. That could change the landscape for memory suppliers like Micron.
This article breaks down the case for Micron as the next Nvidia, the risks that come with it, and what investors should watch for.
The $100 Billion Bet: Micron’s Locked-In Sales for AI
The biggest reason for the excitement around Micron is the $100 billion in sales that the company has already secured. That is not a forecast or a hope. It is a concrete number based on contracts with customers, mostly for a type of memory called high-bandwidth memory, or HBM.
HBM is a special kind of memory that sits very close to the processor in AI systems. It moves data much faster than regular memory, which is critical for training and running large AI models. Nvidia’s GPUs use HBM, and so do the AI accelerators from AMD and Intel. As more companies build out AI infrastructure, the demand for HBM is exploding.
Micron is one of only three companies in the world that can make HBM at scale. The other two are Samsung and SK Hynix, both based in South Korea. That gives Micron a unique position as the only US-based supplier of a key AI component. For customers who worry about supply chain security, that is a big plus.
The $100 billion figure is eye-catching, but it is important to put it in context. In the past, memory companies have announced huge sales commitments only to see those commitments shrink when demand dropped. The question is whether this time is different.
“The demand for AI memory is real and it is growing fast,” says a Wall Street analyst cited by multiple news reports. “But memory markets have a history of overbuilding and then crashing. The question is whether the AI boom is big enough to break that pattern.”
Is the Memory Chip Boom-Bust Cycle Really Dead?
For decades, the memory chip industry has been a textbook example of a cyclical business. When demand is high, prices rise, and companies build new factories to meet that demand. But building a memory factory takes two to three years and costs billions of dollars. By the time the new supply comes online, demand often has slowed. Prices then crash, and the cycle starts over.
This pattern has wiped out companies before. In the 2000s, a glut of DRAM memory led to massive losses across the industry. More recently, in 2019, memory prices fell sharply as smartphone and PC demand slowed. That cycle was brutal for Micron and its rivals.
Some analysts now argue that the AI boom is different. They say that AI demand for memory will keep growing for years, not months. They also point out that building new memory factories is so expensive that only a few companies can afford to do it. That limits the risk of oversupply.
But not everyone is convinced. “The memory industry has seen many ‘new eras’ that turned out to be just another cycle,” a Harvard chip expert told Fortune in a recent interview. “This time may be different, but we have heard that before.”
The Harvard expert, who has studied the chip industry for decades, notes that the current boom is being driven by a single application: AI. If AI investment slows, memory demand could drop sharply. And because memory is a commodity, prices can fall faster than demand.
The Changing of the Guard in AI Hardware: Intel and AMD Rise
While the focus is on Micron, a broader shift is happening in the AI chip market. For the past two years, Nvidia has dominated AI computing. Its GPUs have been the gold standard for training large models, and the company’s market cap hit $1 trillion in record time. But that dominance is starting to show cracks.
CNBC recently reported that Wall Street sees a “changing of the guard in AI.” Intel and AMD shares have been soaring while Nvidia’s stock has lagged. The reason is that companies like Meta, Microsoft, and Google are looking for alternatives to Nvidia’s expensive GPUs. They want chips that are cheaper, more available, or better suited to specific tasks.
Intel has been pushing its Gaudi AI accelerators, which compete directly with Nvidia’s offerings. AMD has its MI300 series, which has won design wins from major cloud providers. Both companies are investing heavily in AI, and their memory needs are just as large as Nvidia’s. That is good news for Micron, because it means more customers for HBM.
But there is a flip side. If Intel and AMD take market share from Nvidia, the overall demand for HBM may not grow as fast as some expect. Nvidia’s GPUs use more HBM per chip than most competitors. A shift to Intel or AMD could mean lower HBM demand per server, even if the total number of AI servers goes up.
“The AI market is broadening, not shrinking,” says a semiconductor analyst quoted in the CNBC report. “But the mix of chips will change, and that matters for memory suppliers.”
Lessons from Nvidia’s Meteoric Rise and Micron’s Potential
Nvidia’s rise to a $1 trillion market cap was unprecedented in the history of the stock market. It took the company just a few years to go from a niche gaming chip maker to the most important company in AI. Along the way, its stock price rose more than 10 times.
Wall Street is now looking for the next Nvidia, and Micron is a natural candidate. The company is a key supplier to the AI industry, it has a strong US manufacturing base, and its stock is still much cheaper than Nvidia’s. The Motley Fool recently noted that Wall Street expects big things from Micron, with some analysts predicting that the stock could double or triple in the next few years.
But comparing any company to Nvidia is dangerous. Nvidia’s success was built on a perfect storm: a massive shift to AI computing, a dominant hardware platform, and a software ecosystem that locked in customers. Micron does not have that kind of lock-in. Its HBM chips are a critical component, but they are also a commodity. If customers can get cheaper HBM from Samsung or SK Hynix, they will switch.
Nvidia also benefited from being the first mover. It started investing in AI chips years before the boom. Micron is a latecomer to the AI story, and it is playing catch-up in HBM production. The company has said it expects to be a leader in HBM by 2025, but it is not there yet.
Harvard’s Warning: ‘This Too Will Pass’ on AI Memory Booms
The most sobering perspective comes from a Harvard chip expert who spoke with Fortune. The expert, whose identity was not disclosed in the report, has a long track record of analyzing the semiconductor industry. In the interview, he warned investors not to get carried away by the current AI memory boom.
“This too will pass,” he said, echoing a famous phrase used to remind people that all booms eventually end. “The memory industry has seen this before. A new application comes along, everyone gets excited, and then the cycle turns.”
The expert pointed to several historical examples. In the 1990s, the PC boom drove huge demand for DRAM. In the 2000s, it was smartphones. In the 2010s, it was cloud computing. Each time, memory companies built new factories, prices crashed, and many players went bankrupt.
“AI is a big deal, no question,” he said. “But it is not the first big deal for memory, and it will not be the last. The fundamentals of the industry have not changed. It is still a commodity business with high fixed costs and volatile demand.”
The expert also noted that the current boom is being fueled partly by government subsidies and tax incentives for chip manufacturing. That could lead to overbuilding, as companies rush to claim subsidies before they run out. “When the subsidies end, we may see a reckoning,” he said.
What Makes Micron Different This Time in the AI Memory Market?
Despite the historical warnings, there are reasons to think that Micron’s position is stronger than in past cycles. The company has learned from its mistakes. After the 2019 downturn, Micron cut costs, reduced debt, and focused on high-value products like HBM. It also invested heavily in research and development, aiming to stay ahead of competitors.
The biggest difference, however, is the nature of AI demand. Training a large AI model requires enormous amounts of memory. A single Nvidia H100 GPU uses 80 gigabytes of HBM. A cluster of thousands of GPUs can use petabytes of memory. And as models get bigger, memory requirements grow.
“AI is not a one-time upgrade,” says a Wall Street analyst cited in the TechCrunch report. “It is a continuous investment that will last for years. That gives memory companies a visibility they have never had before.”
Micron’s $100 billion in locked-in sales is evidence of that visibility. Customers are signing multiyear contracts, which is unusual in the memory industry. In the past, most sales were spot purchases. Today, cloud providers and AI companies are willing to commit to long-term deals because they need guaranteed supply.
Another factor is the consolidation of the memory industry. There are only three major HBM suppliers left: Micron, Samsung, and SK Hynix. That gives them pricing power that smaller players never had. If demand stays high, prices could remain elevated for longer.
The Bottom Line for AI Memory Investors: Micron vs. Nvidia
So, is Micron the next Nvidia? The answer is probably no, at least not in the same way. Nvidia’s rise was driven by a unique combination of hardware, software, and timing that is unlikely to repeat. Micron is a different kind of company, with a different business model and a different set of risks.
But that does not mean Micron is a bad investment. The company is in the right place at the right time. AI demand for memory is real, and Micron is one of the few companies that can meet that demand. The $100 billion in sales is a strong signal that customers are betting on Micron’s future.
The key is to keep expectations realistic. Memory is a cyclical business, and cycles do not die easily. The Harvard expert’s warning that “this too will pass” is a reminder that no boom lasts forever. Investors who buy Micron at today’s prices should be prepared for volatility.
For those who want to ride the AI wave without chasing Nvidia’s high valuation, Micron offers a different path. It is a bet on the infrastructure of AI, not the flashiest part of it. That may be less exciting, but it could also be more sustainable.
As one analyst put it: “Nvidia is the star quarterback. Micron is the offensive line. You need both to win, but the offensive line does not get the glory. And it also does not get injured as easily.”
The bottom line is this: Micron has a real opportunity, but history says to be careful. The smart money will watch the numbers, listen to the warnings, and not get swept up in the hype. Because in the memory business, what goes up can come down just as fast.
Frequently Asked Questions
Why is Wall Street looking at Micron as the next Nvidia?
Wall Street is interested in Micron because it is a key supplier of high-bandwidth memory (HBM), a critical component for AI systems. The company has secured substantial future sales, making it an attractive investment for those seeking to capitalize on the AI boom.
What is high-bandwidth memory (HBM) and why is it important for AI?
HBM is a specialized type of memory that sits very close to the processor, allowing for much faster data transfer. This speed is essential for training and running large AI models efficiently, making HBM a critical component in AI hardware.
What are the historical risks associated with memory chip companies like Micron?
The memory chip industry is known for its cyclical nature. Periods of high demand and prices often lead to overbuilding of manufacturing capacity. When demand eventually slows, this oversupply can cause prices to crash, leading to significant losses for companies in the sector.
What is the warning from the Harvard chip expert about the AI memory boom?
A Harvard chip expert warns that the current AI memory boom may follow historical patterns and eventually pass. They emphasize that the memory industry has experienced similar 'new eras' before, which ultimately turned out to be just another cycle.
How is the competition in the AI chip market changing?
While Nvidia has dominated AI computing, competitors like Intel and AMD are gaining ground with their own AI accelerators. This shift could impact the overall demand for specific types of memory, even as the AI market broadens.
Why is Micron's $100 billion in sales significant?
The $100 billion figure represents concrete contracts for future sales, not just forecasts. This level of commitment from customers is unusual in the memory industry and suggests a strong, ongoing demand for Micron's HBM products.
What makes Micron's current situation potentially different from past memory cycles?
Micron has learned from past downturns, focusing on cost reduction and high-value products like HBM. Additionally, the nature of AI demand is seen as more continuous than previous upgrades, and the consolidation of HBM suppliers gives them more pricing power.
