Memory chip prices surge as data-center boom fuels inflation fears

A surge in demand for memory chips used in AI data centers is driving up prices across the semiconductor industry, adding a fresh source of inflationary pressure that threatens to complicate the Federal Reserve's interest-rate path just as investors brace for a pivotal reading on consumer prices.

"Memory pricing has entered a super-cycle driven by AI infrastructure buildout that shows no signs of abating," Dan Ives, head of technology research at Wedbush Securities, said. "But the question investors are asking is whether these price increases are sustainable or whether they represent a bubble in capital expenditure that will eventually correct."

Micron Technology Inc., the Idaho-based memory chip maker whose products are essential for AI data centers, reported fiscal third-quarter earnings Wednesday that beat analyst estimates, sending shares up 16% in after-hours trade. The company flagged robust demand from AI and data center customers and forecast stronger-than-expected guidance for the current quarter, helping dispel concerns that had triggered a 13% selloff in its stock just two days earlier. Micron's market capitalization has reached $1 trillion, with shares up more than 800% this year through last week's close.

The memory chip price surge represents what some economists describe as a third wave of inflation — one driven not by energy costs or supply-chain disruptions but by the physical infrastructure of artificial intelligence. Globally, AI spending is projected to reach $2.59 trillion in 2026, up 47% year over year, according to research firm IDC. Memory and storage vendors have significantly outperformed the S&P 500 and the SOX semiconductor index since the start of 2025, BNP Paribas data shows.

Why Memory Prices Are Rising

Memory chips — specifically high-bandwidth memory (HBM, a type of DRAM stacked vertically to maximize data transfer speeds) — have become one of the most constrained components in the AI supply chain. Nvidia Corp.'s H100 and Blackwell graphics processing units require HBM3 memory from suppliers including Micron, SK Hynix Inc. and Samsung Electronics Co. Each H100 GPU uses about 80 gigabytes of HBM, and demand has outstripped supply for more than a year.

The tightness in memory supply has pushed prices higher. SK Hynix, which controls roughly half the HBM market, saw its shares slump 12% on June 23 after a South Korean media report said the company was slowing expansion of AI memory chip production and shifting emphasis to commodity DRAM — a move analysts interpreted as a signal that pricing may be peaking. The stock rebounded 8% on June 25 after Micron's strong earnings restored confidence in the demand outlook.

"Any indication of a slowdown in demand for AI is seen as a potential turn in the cycle," Gil Luria, head of technology research at D.A. Davidson, said. "While the overwhelming sense is that demand is still far exceeding supply, investors are waiting for Micron to indicate that is still the case."

The Inflation Connection

The memory chip price increases are feeding into broader inflation measures at a delicate moment for the Fed. The Personal Consumption Expenditures price index — the central bank's preferred inflation gauge — rose 3.8% year over year in April, well above the 2% target and the largest increase in three years. The May reading, due Thursday, is expected to show core PCE at 3.4%, according to consensus estimates.

Fed Chair Kevin Warsh has signaled the central bank may raise its benchmark rate by a quarter percentage point before 2027, with markets pricing a 98% chance of a hike by the September meeting. Higher memory chip prices raise the cost of data-center construction and operation, potentially feeding into services inflation as cloud providers pass costs to customers.

Qualcomm Inc. added to the bullish narrative Wednesday, forecasting $15 billion in sales from its data center business by 2029, sending its shares up 13.3% in after-hours trade. The chip designer's outlook reinforced expectations that semiconductor companies will emerge as the clearest winners of the AI boom, even as the broader economic impact of their success creates headwinds for rate-sensitive growth stocks.

For investors, the tension is between demand tailwinds and macro headwinds. Nvidia trades at about 35 times forward earnings, while Micron's rally has pushed its valuation to levels that leave little room for disappointment. The PCE print Thursday will determine whether the market can sustain both AI optimism and rate-hike expectations simultaneously — a combination that has historically proved difficult to maintain.

This article is for informational purposes only and does not constitute investment advice.