Sanjay Mehrotra, chief executive officer (CEO) of Micron Technology, pointed to customer price-cut pressure as the cause of the memory supply shortage that emerged after the spread of artificial intelligence (AI). He argued that some large customers spurred past declines in memory prices, which weakened the industry's capacity to invest in production facilities.
On the 30th (local time), CEO Mehrotra appeared on CNBC's Mad Money and, regarding the recent memory supply-demand imbalance, said, "Some customers pushed industry prices down excessively."
He cited the deterioration of the memory market in 2023 as the direct backdrop for the pullback in investment. CEO Mehrotra said, "In 2023, memory prices fell to one-third of their previous levels," and noted, "In a situation where corporations were taking losses, they could not invest in new production capacity."
Micron's profitability also deteriorated significantly at the time. Micron's gross margin for fiscal 2023 was -7.3%. Capital expenditures (CAPEX) over the same period were $7.7 billion, down from $12.1 billion a year earlier. CEO Mehrotra said, "It had a serious impact on the industry's ability to invest."
Memory demand is rising rapidly along with the spread of AI servers. In particular, as supply of memory for AI semiconductors centered on high-bandwidth memory (HBM) has tightened, upward price pressure is growing across the DRAM and NAND markets.
CEO Mehrotra said the shortage will be difficult to resolve in the short term. He said, "It takes years to build new semiconductor fabs, and next-generation memory manufacturing processes have become far more complex," noting the memory supply crunch could continue beyond 2027.
Micron is investing about $200 billion in U.S. manufacturing facilities and research and development (R&D) to expand capacity. The company is building new memory plants in Boise, Idaho, and Syracuse, New York.
Rising memory prices are also affecting consumer product prices. CNBC reported that Apple raised prices on some Mac and iPad models, citing higher memory and storage costs. As investment in AI data centers increases, not only server memory but also component prices that go into consumer electronics are under pressure, it said.
The remarks came as the gap widened between suppliers and customers over who bears responsibility for higher memory prices. CEO Mehrotra said the cause of the shortage cannot be seen simply as insufficient production by memory companies and argued that the investment pullback accumulated during past price negotiations must also be considered.