Shenzhen Zhilian Weichuang Technology Co., Ltd.

Will the AI "storage shortage" worsen? Micron President: The supply and demand situation in 2027 will be tighter than this year

Publication Date: 2026-11-20 Source:Cailian Press Key words: AI memory chips HBM View count:

November 20 — Artificial intelligence (AI) demand continues to expand, but memory chip supply is lagging behind the growth pace. Micron Technology executives have recently stated that supply shortages are unlikely to ease in the short term, and the supply-demand imbalance in 2027 and 2028 may be even more severe than in 2026.


Micron Technology President and Chief Technology and Product Officer Scott DeBoer said in an interview that the supply and demand for memory chips will not be balanced for a long time. He pointed out that in the foreseeable future, supply and demand will not find a clear balance, and the supply-demand situation in 2027 is expected to be "tighter than in 2026."


On the 18th local time, Micron Technology CEO Sanjay Mehrotra also stated at the TomorrowX Summit investor conference in Austin, Texas, that 75% of the company's expected 2027 capacity has already been pre-sold. Micron is currently negotiating supply agreements with customers for 2028, but there is still no clear timetable for when supply and demand will return to balance.


He said, "Demand for our business remains strong, and supply remains tight. ”


AI agents drive demand


The core driving force behind this round of storage demand growth comes from the continuous expansion of AI model scale and the rise of AI agents. DeBoer mentioned that AI agents are generating massive demand for computing power and storage, and the growth momentum shows no signs of slowing down. Major chip manufacturers are competing to build higher-performance AI systems, simultaneously driving demand for products like dynamic random access memory (DRAM). Micron has become a beneficiary of this round of demand expansion.


Among them, High Bandwidth Memory (HBM) can transmit data to AI processors at high speed, reducing data latency. Storage capacity and transfer speed have thus become key indicators affecting the computational efficiency of AI systems.


However, tight supply and rising prices are gradually putting cost pressure on downstream customers. Laptop prices have risen due to the shortage of memory chips; Nvidia's interim CEO Colette Kress also mentioned during the Q3 2027 earnings call on the 17th that rising storage costs will erode the company's gross margins.


Demand exceeds Micron's supply capacity


Facing a long-term supply crunch, customers have been locking in forward futures sources in advance. Mehrotra revealed that Micron has signed 26 multi-year customer agreements, involving upfront payments and funding commitments totaling $32 billion; The supply volume customers want over the next five years has already exceeded the scale Micron can promise.


The market is also watching whether customers will seek to terminate contracts if AI demand cools in the future. Mehrotra emphasized that these contracts were signed after thorough consultation between the client's management and the procurement department, and include strict purchase or payment obligations.


"I don't think clients will easily tear up contracts." He emphasized.


To expand supply, Micron is advancing a U.S. investment plan totaling over $250 billion, aimed at increasing DRAM capacity while also expanding existing fabs worldwide.


DeBoer admits that building new capacity "takes a long time," which is one of the supply challenges the company is working to alleviate. Moreover, even with ongoing investment and expansion, new capacity still struggles to meet rapidly growing demand immediately.