Micron CEO says AI memory demand has permanently transformed the boom-and-bust memory industry

AI Transforms Memory From Commodity to Strategic Infrastructure

Micron Technology CEO Sanjay Mehrotra declared that artificial intelligence has fundamentally altered the memory industry, ending decades of boom-and-bust cycles that historically plagued memory chip manufacturers1

1

. Speaking on CNBC's "Mad Money," Mehrotra emphasized that "today there is no AI without memory" and explained how AI systems require more memory, higher performance memory, and lower power memory than traditional computing applications1

1

. This shift has transformed memory from a commodity component purchased primarily on price into what Mehrotra calls the "strategic infrastructure of the AI era"3

3

.

The memory industry has historically experienced cyclical patterns where strong demand encouraged manufacturers to add capacity, only for excess supply to eventually drive down prices. However, AI-driven demand for memory chips appears to be creating a more durable demand environment. Mehrotra, an engineer with over 40 years in the chip industry who previously co-founded SanDisk, believes this transformation represents a permanent structural shift rather than another temporary cycle1

1

.

Supply Crunch Shows No Signs of Easing

Micron Technology currently faces unprecedented demand that far exceeds its production capacity. Mehrotra revealed that data centers are requesting roughly 50% more supply than Micron can commit, and emphasized that "all our customers across our end markets will buy everything that we make"1

1

3

3

. The CEO stated bluntly: "We see no end when supply catches up with demand, and the demand continues to grow"2

2

.

Yorkville Ives Senior Managing Director Dan Ives told CNBC that the memory trade has become "foundational, not cyclical," estimating demand-to-supply ratios in AI memory at roughly 15-to-12

2

. He added that "memory players, right now it's their world and everyone else is paying rent"2

2

. Daniel Newman, CEO of The Futurum Group, suggested that Mehrotra "finally said the quiet part out loud," noting the AI-driven demand cycle "may never end" as data centers give way to robotics, autonomous vehicles, and physical AI applications2

2

.

Massive U.S. Manufacturing Investment in Boise, Idaho

Micron Technology is backing its optimistic outlook with a planned $250 billion investment in U.S. manufacturing and research1

1

. The company is constructing a massive semiconductor fabrication site near its headquarters in Boise, Idaho, that will eventually include two fabrication plants, each roughly the size of 10 football fields1

1

. According to Mehrotra, a single fab will contain enough steel rebar to circle Earth twice1

1

. The first Boise fab is expected to begin producing wafers in mid-20271

1

3

3

.

Additionally, Micron disclosed the launch of Micron Research Labs, a $10 billion, 10-year investment in a U.S.-based research hub anchored by a flagship campus in Boise, aimed at advancing memory and AI technologies2

2

. The scale of these investments reflects how dramatically Mehrotra believes AI has altered the outlook for the memory industry1

1

.

Long-Term Customer Agreements Reduce Market Volatility

Micron Technology is gaining greater visibility into future demand through long-term customer agreements, representing another important shift for a business historically exposed to swings in the spot market. During the company's most recent earnings call in late June, Micron announced it had signed five-year strategic agreements with 16 customers, and Mehrotra confirmed the company has since inked additional deals1

1

3

3

. These agreements, which can be extended, commit customers to taking supply and give Micron "assurance of demand"1

1

2

2

.

The nature of customer relationships is also evolving. Rather than customers simply soliciting bids and buying from whichever supplier offers the lowest price, high-performance memory for AI systems must increasingly be designed alongside the processors and systems in which it will operate1

1

. Mehrotra explained that "we are working closely with them earlier and earlier in their development cycle" and that customers "recognize the value of memory, because memory is what is enabling them to design products that are driving growth engines for them"1

1

. This shift makes memory a strategic infrastructure component essential to system performance rather than simply another interchangeable part1

1

.

Supply Relief Unlikely Before 2028

Even as Samsung, SK Hynix, and Micron Technology all expand capacity, meaningful production relief is unlikely before 2028 at the earliest, according to Counterpoint Research Director MS Hwang, since AI demand is now pulling in conventional DRAM alongside high-bandwidth memory2

2

. Mehrotra made a similar point in June, telling investors on Micron's fiscal third-quarter earnings call that "even as we expect industry supply to improve gradually in 2028, we currently do not have line of sight as to when memory supply will be able to catch up with increasing demand"2

2

.

The opportunity extends beyond data centers. Mehrotra expects autonomous vehicles, robots, and AI-enabled consumer devices to require increasingly large amounts of memory in the years ahead1

1

3

3

. This expanding addressable market suggests demand could continue growing even as data center buildouts mature.

However, not everyone agrees the memory industry's transformation is permanent. ARK Invest founder Cathie Wood has taken the opposite view, avoiding memory suppliers like Micron and SK Hynix as today's extraordinary pricing is more likely to invite competition and engineering workarounds than persist indefinitely2

2

. Mehrotra acknowledged this uncertainty, noting that "nothing ever goes in a straight line"2

2

, though he maintains confidence in the structural nature of AI-driven demand for memory chips.