only.Nvidia’s Shift and AI Chip Shortages May Drive Gadget Prices Higher

Analysts warn that the price of your next smartphone could climb as the AI boom tightens semiconductor supply chains. A recent product‑line change by Nvidia may aggravate the situation.

Globally, tech giants are pouring hundreds of billions of dollars into AI data‑center infrastructure. Those data centers rely on high‑performance GPUs and specialized memory chips supplied by companies such as Nvidia, which in turn depend on a complex network of component manufacturers.

Other key players—AMD, hyperscalers like Google and Microsoft, and a host of memory‑module makers—share the same supply chain. When any link in that chain falters, it reverberates across the entire ecosystem of consumer electronics.

Supply bottlenecks are already pushing up the prices of critical components, raising the specter of higher retail prices and even shortages of flagship devices.

“The surge in AI‑driven data‑center demand is creating bottlenecks across multiple segments of the semiconductor supply chain,” said Peter Hanbury, partner in the technology practice at Bain & Company.

Alibaba’s chief executive Eddie Wu highlighted shortages across semiconductor fabs, memory‑chip producers, and storage‑device manufacturers. Alibaba is building its own AI infrastructure and designing custom chips, giving Wu a front‑row view of the constraints.

“We are seeing a pronounced undersupply that will act as a major bottleneck for the next two to three years,” Wu warned.

Hanbury added that hard‑disk drives (HDDs), still favored by hyperscalers for bulk storage, are at capacity. To meet demand, many firms have pivoted to solid‑state drives (SSDs), which are also essential for smartphones, tablets, and laptops.

Dynamic random‑access memory (DRAM), especially high‑bandwidth memory (HBM) used in Nvidia’s GPUs, is another pressure point. The race to secure HBM stacks is crowding out other memory products, driving up prices industry‑wide.

Expanding semiconductor capacity is capital‑intensive and time‑consuming. Foundries and component makers, wary of over‑investing in a market they deemed overly optimistic, did not build the extra fab lines that today’s demand requires.

“The direct cause of the shortage is the rapid increase in demand for data‑center chips,” Hanbury explained. “Suppliers hesitated to over‑build, and now adding new capacity takes two to three years.”

Nvidia dominates the AI‑data‑center GPU market and is a major customer for high‑bandwidth memory produced by TSMC. Recently, Nvidia has accelerated its shift toward low‑power double‑data‑rate (LPDDR) memory—a staple of premium smartphones from Samsung and Apple.

Because Nvidia’s orders are on the same scale as the world’s largest smartphone manufacturers, its growing appetite for LPDDR is adding unprecedented pressure on a segment of the memory market that was already tight.

“Nvidia’s pivot to LPDDR creates a seismic shift for the supply chain. The market cannot easily absorb a demand surge of this magnitude,” said Minsoo Hwang, research director at Counterpoint Research.

The ripple effect is clear: when foundries prioritize AI‑centric chips, other semiconductors—especially DRAM and SSDs—receive less allocation. Those parts constitute roughly 10 % to 25 % of a typical smartphone or PC’s bill of materials.

Counterpoint Research projects memory‑chip prices will rise 30 % in the fourth quarter of 2025 and an additional 20 % in early 2026. A 20 %–30 % increase in DRAM and storage costs could lift total device BOM costs by 5 %–10 %.

“Component costs are already climbing, and we expect the impact to accelerate into next year,” Hanbury noted.

Industry leaders are sounding the alarm. Xiaomi, the world’s third‑largest smartphone maker, warned of “a sizeable rise in product retail prices.” Dell’s chief operating officer Jeff Clark described the component‑price surge as “unprecedented,” noting rapid cost increases across memory and storage categories.

AI data‑center demand is also crowding out legacy chips used in automotive, industrial, aerospace, and defense applications. As manufacturers reallocate capacity to serve AI customers, these downstream industries may face higher component costs and delayed product launches.

“Many markets depend on the same manufacturing capabilities as the data‑center sector. Price pressures will likely cascade across multiple verticals,” Hanbury concluded.