Product1 publisher3 min readPublished
Memory diverted to AI data centers is lifting phone prices in India four times faster than in the US
Chinese phone makers are cutting entry-level models as AI data centers absorb memory chips, and phone prices are up 21% in India against 5% in the US. Teams counting on first-time phone buyers outside the US now have to plan around a more expensive entry device.
The Product Desk · Product desk
Drafted by a language model from the sources cited here and checked against its claim ledger before publication. How we use AISend a correction
What happened
- Prices of existing smartphone models have risen about 15% globally this year, and newly launched models cost roughly 25% more than last year's, Rest of World reported.
- IDC's Ramon Llamas said all three main memory makers moved most of their supply to AI data centers in late 2025, leaving many companies scrambling for inventory.
- Omdia data show that in Southeast Asia, Oppo shipped 96% fewer sub-$100 phones, and Vivo pushed the price of its main entry-level model past $100 in most markets.
- In Africa, 81% of phones shipped last year cost under $200. In the second quarter of 2026, shipments there of phones under $100 were down 34% from a year earlier.
Compiled by The Product DeskSomething wrong?How this is made
Why it matters
- exposure Fewer affordable first phones in India, Asia-Pacific and Africa put the GSMA's forecast of nearly 800 million new mobile internet users by 2030 at risk, by the group's own account.
- constraint Services that count on new users arriving with a sub-$150 phone will find fewer of those phones on sale in the markets where they planned to grow.
- precedent If costs stay above pre-2025 levels, as Lam expects, the cheapest phone tier resets higher for good, and device-price assumptions written before 2025 need replacing.
In December, Xiaomi launched the 128GB Redmi 15C in India at 12,499 rupees. By June the same phone cost 16,999 rupees, a 36% increase [5]. A buyer who waited six months paid 4,500 rupees more for the same model [1].
Teams' growth plans assume the next user in India buys a cheap smartphone and installs the app. Last year's market supported that assumption. More than one in four smartphones shipped worldwide in 2025 cost less than $150 [6]. Chinese makers including Xiaomi, Oppo, Vivo and Huawei built their businesses on affordable phones [8], and Chinese brands account for about 60% of global shipments [7].
Ivan Lam, a senior analyst at Counterpoint, told Rest of World that Chinese makers have "drastically reduced" entry-level projects this year as memory costs rose [2]. The companies "would rather reduce investment in products below $150 or $200 and put more resources into higher-margin phones, branding, user experience, and AI," he said [9].
The cost lands unevenly. Phone prices are up 19% across Asia-Pacific and 18% in the Middle East and Africa [10]. India's increase is a little over four times the US one [2].
How long this lasts decides how much replanning it deserves. Lam said he does not expect costs to return to pre-2025 levels [11]. "What used to be below $150 may become below $250, or even $300," he said [12]. Ramon Llamas, research director in mobile devices at IDC, told Rest of World the shortage is unlikely to ease soon [13]. Samsung Electronics, SK Hynix and Micron Technology make more than 90% of the world's memory chips [19]. Demand from the data center side keeps climbing: S&P Global estimates US hyperscaler capital expenditure at $470 billion in 2025, a projected $870 billion in 2026 and more than $1.3 trillion in 2027 [14]. That is close to triple in two years [3].
Operators building phone-first services for these markets face the same limit the GSMA describes. "Stakeholders are increasingly using AI to provide access to important services, but people won't be able to benefit without access to an internet-enabled phone and the internet," said Claire Sibthorpe, head of digital inclusion at the GSMA's nonprofit foundation [15].
For a team deciding what to do on Monday, I'd use a 2x2. One axis is where the next users come from: people buying a phone, or people who already own one. The other is the market: India, Asia-Pacific or the Middle East and Africa, with increases of 18% to 21%, or the US at 5% [10]. New buyers in the high-increase markets are the exposed quadrant. Existing owners there are safer for now, though if replacements cost more, the handset an app has to support gets older each year. US users in either quadrant face a single-digit increase [10].
In the exposed quadrant, I'd plan acquisition around phones already in people's hands and slow the forecast for first-time users. The tradeoff is testing time on older handsets and a ceiling on features, and that effort is wasted if memory prices fall faster than Lam and Llamas expect. The one assumption in the plan to update is the device price the next cohort is expected to afford. If that figure sits in the old sub-$150 tier, Lam's forecast puts the new entry tier $100 to $150 higher [4].
What to watch
- Whether Samsung, SK Hynix or Micron move any memory supply back toward consumer electronics as US hyperscaler capex heads past $1.3 trillion in 2027.
- The next quarter of sub-$100 shipment data for Africa and Southeast Asia, to see whether the 34% and 96% declines deepen or level off.
- Any GSMA revision of its forecast of nearly 800 million new mobile internet users by 2030.