---
title: "Thanks for the memory: how Micron became an unlikely kingmaker in the AI age"
publisher: "Stockmark.IT"
author: "Stockmark.IT Website"
published: "2026-06-29T06:45:52+00:00"
modified: "2026-06-29T06:45:52+00:00"
date: 2026-06-29
canonical: "https://stockmark.it/thanks-for-the-memory-how-micron-became-an-unlikely-kingmaker-in-the-ai-age/"
category: "Business"
categories: ["Business", "Technology"]
image: "https://stockmark.it/wp-content/uploads/2026/06/stencil.default-2026-06-29T074316.316.avif"
format: "news"
language: "en-GB"
---

# Thanks for the memory: how Micron became an unlikely kingmaker in the AI age

**Published:** June 29, 2026
**Author:** Stockmark.IT Website
**Categories:** Business, Technology
**Featured image:** ![Micron employee in cleanroom operating semiconductor manufacturing equipment. from Stockmark.it](https://stockmark.it/wp-content/uploads/2026/06/stencil.default-2026-06-29T074316.316.avif)

---

When Micron Technology was founded in 1978, in basement offices beneath a dentist’s practice in Boise, Idaho, few would have imagined that one day the company’s fortunes would help determine the mood of global markets. Yet that is precisely where Micron finds itself now. A business once associated with the grind of commodity memory chips has become one of the clearest beneficiaries of the artificial intelligence boom, propelled by a specialised product whose scarcity has turned it into one of the most prized components in modern computing.

This year the company’s shares have risen by more than 300 per cent, lifting its market value beyond $1 trillion and placing it among the world’s most valuable corporations. For investors searching for a way to participate in the AI surge without buying the obvious names, Micron has become a proxy for something larger: the idea that the infrastructure behind AI may prove as lucrative, and perhaps more durable, than the software models that dominate the headlines.

The company’s latest quarterly results only sharpened that view. Revenue more than quadrupled to $41.5 billion, while profit rose nearly 15-fold to $28.2 billion. Sanjay Mehrotra, Micron’s chief executive, said those numbers reflected the “strategic value of memory in the AI era”. In the language of corporate earnings calls, that is not a throwaway line. It is a statement about a structural shift in the semiconductor market. What was once a cyclical, low-margin industry, vulnerable to boom and bust, has been recast by the demands of AI systems that require vast quantities of fast, reliable memory to function at scale.

The market’s reaction was immediate. Micron’s shares climbed a further 16 per cent the next day, adding about $200 billion to its market capitalisation. On a day when tech stocks were otherwise selling off across the globe, Micron’s results were treated as a bellwether for the AI trade as a whole. Analysts at Cboe captured the mood before the announcement, writing that “all eyes are on Micron’s earnings”. They were not wrong. The company has become a barometer for investor confidence in the long-term economics of AI, precisely because its memory chips sit at a crucial point in the supply chain rather than at its most glamorous end.

Micron’s rise is all the more striking given the circumstances of its birth. The founders, twin brothers Joe and Ward Parkinson, along with Dennis Wilson and Doug Pitman, began as engineers at Mostek, one of the early memory chip pioneers. Their first office was so unassuming that Joe Parkinson later joked about the “happy gas” drifting down from the dentist upstairs, saying the arrangement probably “knocked our IQ down about ten points”. The humour masks a more serious truth about American semiconductors in the late 1970s. New entrants needed ingenuity, but they also needed capital, patience and a willingness to operate on the edge of a brutally competitive market.

At first, Micron was little more than a design consultancy, producing chip layouts for other manufacturers. Within a year, however, the company decided to manufacture its own chips, a move that required a fabrication plant and a willingness to compete against far larger rivals. This was no trivial gamble. Even then, building a fab was ruinously expensive, and the company entered the market at a moment when Japanese producers were flooding it with inexpensive semiconductors. Yet Micron found advantages where others saw only obstacles. It secured $300,000 in local backing over lunch, followed by a $1 million investment from JR Simplot, the Idaho potato magnate whose fortune had been made supplying McDonald’s with fries. Idaho’s cheaper land and lower energy costs helped Micron build at a fraction of the cost faced by competitors, while the company’s engineers designed a dynamic random access memory chip that was smaller than anything else available, allowing more chips to be produced from each silicon wafer.

That technical edge mattered because Dram, the short-term memory of a computer, is fundamental to how processors operate. It temporarily stores data that the central processor needs to access immediately. Micron’s smaller, more efficient chips gave it a fighting chance in a market where scale was becoming decisive. Even so, the wider industry was heading into a period of extraordinary contraction. Between 1978 and 1986, the United States’ share of the Dram market fell from 70 per cent to 20 per cent, while Japan’s rose from 30 per cent to 75 per cent. Micron survived where most of its domestic peers did not.

The US government stepped in after Micron petitioned for help, and in 1986 Washington and Tokyo agreed to a semiconductor arrangement that imposed a price floor on Japanese companies. By then, seven of the nine major American Dram producers had either gone bankrupt or abandoned the market. Micron and Texas Instruments remained, with Texas Instruments eventually selling its Dram unit to Micron in 1998. The episode helped reshape the international memory industry, but it also had another consequence. As US buyers looked elsewhere for cheaper supply, South Korean manufacturers rose quickly, and by 1992 Samsung was the world’s largest memory supplier. By 2006 South Korea controlled more than half the global memory market.

For decades, memory remained a difficult business to love. It was highly cyclical and largely commoditised. When demand rose, manufacturers invested heavily in new capacity. By the time those plants were ready, demand often had softened, creating a glut that drove prices down again. Investors learned to treat the sector with caution. Even Micron, despite its technical competence and survival instincts, could not escape the pattern. In 2023 the company reported an annual loss of $5.8 billion after producers had built out too much capacity in response to the pandemic-era surge in consumer electronics demand. When that boom faded, prices collapsed.

The AI revolution has altered that equation. The new centre of gravity in memory is high-bandwidth memory, or HBM, a highly specialised chip that stacks multiple Dram layers to create far greater bandwidth. That makes it possible to move and process large volumes of data at high speed, which is exactly what AI models require. For the first time in years, memory is not just a commodity sold on price. It is a differentiated component, and one that is now indispensable to the largest technology companies in the world.

South Korea’s Samsung and SK Hynix were early leaders in HBM, but Micron closed the gap with its HBM3E design in 2023. The following year it signed a deal with Nvidia to supply memory for the Blackwell generation of AI chips, a contract that confirmed its place in the critical path of the AI supply chain. Demand has since been described as insatiable, driven by the hyperscalers who are spending heavily on the physical infrastructure that underpins artificial intelligence. Yet supply remains constrained. Micron has said its HBM production is sold out through 2026 and that it can currently meet only about 50 to 66 per cent of customer demand.

That mismatch is central to the current investment case. Deutsche Bank analysts have argued that the old boom-and-bust logic may no longer apply in the same way because AI demand is not merely cyclical but structural. In their view, chipmakers are struggling to keep pace with hyperscale demand for memory used in AI, which means the market’s constraints are now the story. Morgan Stanley says Micron, Samsung and SK Hynix together control the entire HBM market. Their dominance has transformed what was once a dull corner of the semiconductor sector into one of its most important battlegrounds.

The effects are spilling beyond the AI sphere. As memory makers redirect production towards HBM, prices for ordinary Dram chips used in consumer electronics have risen sharply. TrendForce estimates that memory chips used in MacBooks increased by 60 per cent in the first quarter of this year alone and are expected to rise by 80 per cent in the second quarter. Apple has already warned that it may need to raise prices on some products by as much as 25 per cent. Tim Cook has described the shortage as “a hundred-year flood”, a striking phrase that reflects both the scale and the unpredictability of the present squeeze.

For Micron, the shortage has become an opportunity. The company is earning record margins not only on premium AI products but also on legacy memory chips, which are suddenly more valuable because of constrained supply. Mehrotra said on the earnings call that he and his team “currently do not have line of sight as to when memory supply will be able to catch up with increasing demand”. In another era, that sort of remark might have sounded like a warning. At present, it reads more like a declaration of strength. Scarcity, in this market, has become a source of power.

The company’s transformation is all the more notable because it reflects a broader reordering of the semiconductor industry. The Philadelphia Semiconductor Index has gained 84 per cent, well ahead of the wider market, but the gains have not been evenly distributed. Samsung’s shares are up 164 per cent this year and SK Hynix’s by 294 per cent, pushing both past $1 trillion in market capitalisation. The same dynamic that has elevated Nvidia has also lifted the memory makers, whose chips are now essential to the AI stack. What they supply is not software, and not even the processing power that headlines usually celebrate, but the capacity to remember, retrieve and move information quickly enough for AI systems to operate at all.

There is, then, a paradox at the heart of Micron’s ascent. A business once defined by commodity logic has found itself at the centre of a market that prizes specialisation and scarcity. The result is a company that can no longer be understood simply as a cyclical chipmaker. It is instead a crucial supplier to an infrastructure build-out that may continue for years, possibly longer, as technology companies race to expand the tools and systems that make AI commercially viable.

That does not mean the old risks have vanished. The memory market has a long history of overexpansion, and every boom tends to breed the conditions for the next downturn. But for now, the balance of power has shifted. Micron’s journey from a basement beneath a dentist’s office to a company with a market value north of $1 trillion says as much about the economics of AI as any keynote speech or product launch. In a business once thought too ordinary to matter, the ability to store and move data has become one of the most valuable capabilities in global technology.

---

**Original URL:** https://stockmark.it/thanks-for-the-memory-how-micron-became-an-unlikely-kingmaker-in-the-ai-age/
*Created by [WP Markdown Endpoint](https://wpmarkdownendpoint.com/)*