Download this article in PDF format.
Artificial intelligence (AI) had a banner year in 2025 as companies moved from experimentation to deployment. AI tools became part of day-to-day operations across industries, increasing spending on data centers, infrastructure and computing hardware.
That demand is now affecting the semiconductor supply chain. As chipmakers prioritize memory used in AI systems, supplies of conventional memory are tightening, with implications for availability, pricing and production planning across consumer electronics, automotive and enterprise markets.
All of this popularity is straining another part of the technology supply chain: memory chip production. As chip manufacturers shift capacity toward higher-margin AI-related products, traditional memory supplies are tightening, costs are rising and lead times are lengthening out.
“An acute global shortage of memory chips is forcing artificial intelligence and consumer-electronics companies to fight for dwindling supplies, as prices soar for the unglamorous but essential components that allow devices to store data,” Reuters reports. It says Japanese electronics stores are limiting the number of hard-disk drives shoppers can buy; Chinese smartphone makers are warning of price increases; and Microsoft, Google and ByteDance are all working ahead to secure supplies from memory-chip makers.
“The squeeze spans almost every type of memory, from flash chips used in USB drives and smartphones to advanced high-bandwidth memory (HBM) that feeds AI chips in data centers,” Reuters adds, noting that prices in some segments have more than doubled since February. "The memory shortage has now graduated from a component-level concern to a macroeconomic risk," Greyhound Research’s Sanchit Vir Gogia, told the publication.
No Silver Bullet
A quick solution to the chip shortage problem seems unlikely at this point. In “AI boom is fueling a memory chip shortage that could hit cars and phones,” CNBC says both chipmakers and analysts are warning of a memory chip shortage that could hit the consumer electronics and automotive industries in 2026, “as companies prioritize massive demand from the artificial intelligence boom.”
Analysts say these supply constraint concerns come as chip manufacturers focus on advanced memory chips used in AI computing, with less focus on production needed for consumer products. “The AI build-out is absolutely eating up a lot of the available chip supply, and 2026 looks to be far bigger than this year in terms of overall demand,” analyst Dan Nystedt told CNBC. The issue is that AI servers run mostly on processors from chip designers like Nvidia, and those processors use a type of memory known as high-bandwidth memory (HBM). The news outlet says the memory industry has also experienced downturns in recent years, and that’s led to under-investment in the industry.
“They’re building new capacity now,” it adds, “but it will take time to get running.”
No Firm Supply Commitments
The potential impacts of the memory chip crisis came to light in November, when the CEO of Semiconductor Manufacturing International Corp. (SMIC) delivered a warning during an earnings call. Manufacturers of cars and smartphones that use memory will face pricing pressure and supply uncertainty next year, he said, and nobody is offering “firm supply commitments” right now.
"People don't dare place too many orders for the first quarter next year," he said, according to Reuters. “Because no one knows how many memory (chips) will actually be available, [or] how many phones, cars, or other products it can support."
