AIがメモリ不足を引き起こし、インドのスマートフォン市場に激震AI-driven memory crunch jolts India’s smartphone market

India's smartphone market is being reshaped by an unexpected side effect of the artificial intelligence boom: a shortage of memory chips that is pushing device prices higher and squeezing manufacturers' margins. The development matters because India is among the world's largest and most price-sensitive smartphone markets, where even small increases in component costs can ripple across hundreds of millions of purchasing decisions.

At the heart of the issue is the growing appetite of on-device AI features for random-access memory. Generative assistants, real-time translation, photo-editing tools and other machine-learning workloads increasingly run directly on handsets rather than in the cloud, and these features demand more memory to operate smoothly. Where budget and mid-range phones once shipped comfortably with 4GB or 6GB of RAM, manufacturers now feel pressure to fit 8GB, 12GB or more to support AI capabilities, raising the amount of memory required per device.

That rising per-unit demand is colliding with tighter global supply. Memory manufacturers such as Samsung, SK Hynix and Micron have been prioritising high-bandwidth memory (HBM), a specialised and highly profitable type of chip used in the data-centre accelerators that power large AI models. Shifting production capacity toward HBM appears to be constraining the availability of the conventional mobile memory, including LPDDR DRAM and NAND flash storage, that smartphones rely on. The result is upward pressure on component prices that eventually reaches consumers.

For India specifically, the effect is likely to be more pronounced than in wealthier markets. A large share of devices sold in the country fall in the entry-level and mid-range segments, where profit margins are thin and buyers are highly sensitive to price. When memory costs rise, manufacturers face a difficult choice: absorb the expense and accept lower margins, pass it on through higher retail prices, or reduce memory configurations and risk offering a weaker experience. Any of these responses can dampen demand or alter the competitive landscape.

Local and international brands operating in India, including Xiaomi, Samsung, Vivo and Oppo, source memory from the same constrained global pool, meaning the pressure is broadly felt. Some manufacturers may respond by staggering launches, adjusting specifications, or concentrating AI features on premium tiers where higher prices are easier to justify. Such moves could slow the spread of on-device AI to the mass market that India represents.

The situation reflects a wider tension in the semiconductor industry, where the economics of AI are pulling resources toward the most lucrative products. Analysts have noted for some time that the build-out of AI data centres, driven by companies training and deploying large language models, is influencing supply and pricing across multiple chip categories. Memory is a commodity market historically prone to cycles of oversupply and shortage, and the current tightening appears to be amplified by the reallocation of manufacturing capacity toward AI-related products.

There is also a policy dimension. India has been pursuing domestic semiconductor and electronics manufacturing through initiatives such as its production-linked incentive schemes, aiming to reduce reliance on imports and build local assembly capacity. However, advanced memory fabrication remains concentrated among a handful of global players, so nearer-term relief for the Indian market is likely to depend on decisions made by those suppliers rather than on domestic production.

For consumers, the practical consequences may include higher prices for a given specification, fewer aggressively priced options at the low end, or devices that offer less memory than buyers might expect at a particular price point. Whether the shortage proves temporary or persistent will depend on how quickly memory makers expand capacity and how demand for both AI hardware and consumer devices evolves in the coming quarters.