AI’s memory chip crunch is crushing the smartphone market

The smartphone market experienced its most significant decline on record in the September quarter, with shipments dropping by 22%, according to market researcher IDC. This downturn was mainly due to the rising costs of memory chips, which surged by 350% over the past year. These memory price hikes were primarily driven by increased competition from AI industry players.

Consequently, smartphone manufacturers faced reduced profits, forcing them to raise prices, trim specifications, and discontinue budget-oriented models. Senior research director for Worldwide Devices at IDC, Nabila Popal, stated that Chinese brands like Xiaomi Corp. and Oppo, which sell lower-priced phones with thinner profit margins, were among the hardest hit.

On the other hand, Apple and Samsung were somewhat insulated from the memory price surge because their phone lineups comprised higher-margin premium devices. A significant portion of the memory chip manufacturers, including Samsung, SK Hynix, and Micron, also produce memory for AI processing. These manufacturers use DRAM and NAND memory chips to create high-bandwidth memory (HBM) chips, which are crucial for AI workloads due to their high data transfer rates.

AI accelerators like Nvidia's graphics processing units (GPUs) demand large quantities of HBM to store model parameters and retain contextual information during problem-solving. HBM is physically embedded within the same package as the GPU, adjacent to its processing silicon. Memory chip manufacturers have been shifting their focus toward HBM production, as generating revenue from this application is more lucrative than manufacturing memory chips for smartphones.

Popal explained that long-term agreements with memory manufacturers, mainly held by AI infrastructure and cloud service providers like Google, Meta, and Amazon, guarantee steady demand for their products through 2030 or 2031. However, memory prices continue to rise as the AI data center expansion is only in its early stages. IDC predicts that worldwide spending on AI infrastructure, including servers and storage, will hit $1.21 trillion annually by 2030.

Memory manufacturers are expanding their capacity, but they aren't rushing into it because they're profiting from inflated memory prices. As a result, smartphone consumers will ultimately bear the costs of this memory crisis. The average selling price of smartphones has already increased considerably, with last year's average being $450, and this year's forecasted figure is $581.