The AI boom Is making smartphones costlier

AI's exploding demand for memory chips is squeezing smartphone supply and pushing prices sharply higher

When Prime Minister Narendra Modi inaugurated SEMICON India 2026 in New Delhi on September 17, the spotlight was on India's ambition to become a serious semiconductor player.

But there is another semiconductor story that deserves attention: the chip shortage created by the AI boom is already reaching the smartphone in your hand.

AI data centers consume enormous quantities of processors and memory. As companies race to build AI data centers, semiconductor manufacturers are prioritizing products for this rapidly growing market. The result is pressure on memory chips used in smartphones (and laptops), and sharply higher prices. Pankaj Gulati, President of CDIL Semiconductors, and a veteran of India's semiconductor and electronics industry, told the author: "AI usage is rapidly increasing, resulting in huge demand in data centers. Consequently, chips are being diverted to data centers, who are also willing to pay higher prices. As availability of chips for smartphone manufacturers gets squeezed, their prices are increasing. Memory is now the single most expensive component in a premium smartphone."

Counterpoint Research confirms this and estimates that the memory component alone accounts for more than 40% of the bill of materials of a premium smartphone.

DRAM and NAND

A smartphone has two types of memory: DRAM and NAND.

DRAM, or Dynamic Random-Access Memory, is the phone's 'worktable' or fast short-term memory. When you open WhatsApp, switch between apps, or play a game, the phone needs fast temporary space in which to work. That is DRAM, more commonly called RAM. It is extremely fast but loses its contents when the phone is switched off.

NAND is more like the phone's 'cupboard' or its non-volatile long-term storage flash memory. It stores contacts, photographs, videos, apps and documents and retains them even when the phone is switched off. So, when a phone is advertised as having 12GB RAM and 256GB storage, the 12GB refers to DRAM and the 256GB to NAND-based storage.

The AI connection

Modern AI Data Centers require enormous quantities of GPUs (Graphics Processing Units or logic chips), ultra-fast HBM (High Bandwidth Memory) or '3D stacked DRAM', and NAND.

Major semiconductor manufacturers are redirecting their production lines toward GPUs, HBM and NAND for massive AI data centers, reducing the capacity available for conventional consumer DRAM and NAND.

This does not mean AI companies are simply taking the exact chips meant for smartphones. Different applications use different types and specifications. The bigger issue is that semiconductor manufacturing capacity cannot be expanded overnight. As manufacturers shift investment and production towards the rapidly growing AI market, the supply-demand balance for conventional memory can tighten. When supply becomes constrained while demand remains strong, prices rise.

Increase in memory prices

Counterpoint says memory prices increased sharply through 2026 and that the resulting cost pressure has substantially reshaped smartphone economics. In Q2 2026, memory prices rose by more than 80% quarter-on-quarter. In premium smartphones, DRAM has become the most expensive individual component.

The effect is now visible in retail prices. Smartphone prices have risen 21% on an average in India in 2026 and by 15% globally. More than 40% of smartphone models have seen price increase.

Manufacturers are responding in several ways. Some are raising prices; others are reducing storage capacities, simplifying camera configurations or introducing more 4G models in selected segments. The pressure is particularly severe in lower-priced smartphones because memory represents a larger share of their overall cost.

Impact in India

For India, this is particularly important. The sub-Rs 15,000 smartphone segment is already under pressure from rising memory and component costs. These affordable devices are not merely communication tools. For millions of Indians, they are gateways to UPI payments, online education, government services, telemedicine and entertainment. There is also an interesting Indian connection to the global memory story.

Micron, one of the world's major memory and storage companies, is headed by Indian-origin Sanjay Mehrotra, its Chairman, President and CEO. Micron's semiconductor assembly and test facility at Sanand in Gujarat began commercial production earlier this year.

This makes SEMICON India particularly relevant. India's semiconductor ambition is not simply about attracting factories. It is about becoming part of a global industry whose products increasingly determine the cost and capability of the digital devices people use every day.

AI promises smarter services, greater productivity and powerful new applications. Yet the infrastructure being built to power that revolution is simultaneously putting enormous pressure on one of the most basic components inside our smartphones.

The semiconductor race, therefore, is no longer just about data centers or chip factories. It may also decide how much the next smartphone in your hand costs.

The author is an alumnus of IIM Ahmedabad and Professor of Practice at IILM University, Gurugram with an interest in AI, Technology and Strategy; Views presented are personal.