Smartphone Shipments Hit 13-Year Low as AI Eats the Memory Supply

A Demand Shock Rooted in the Supply Side

Global smartphone shipments dropped 11% in the second quarter, sinking to levels not seen since 2013, according to early estimates from Counterpoint Research. The proximate cause is a prolonged shortage of memory chips, specifically DRAM and NAND, that has sent handset prices sharply higher and suppressed demand across most price tiers.

This is not a typical inventory correction or a cyclical demand pause. The shortage is structural. Three companies (Samsung, SK Hynix, and Micron) control more than 95% of global DRAM production, and they have systematically reallocated capacity toward high bandwidth memory for AI accelerators. Tech giants are on track to spend roughly $650 billion on AI infrastructure this year, up about 80% from 2025's record. Memory manufacturers are locking in multiyear agreements at premium prices with hyperscalers. Smartphone makers are left competing for whatever remains.

Data centers are now forecast to consume 70% of all memory chips produced globally in 2026, up from 20 to 30% as recently as 2022. That reversal happened fast. It will not unwind soon.

The Bill of Materials Has Become Unworkable

The math at the low end of the market has broken. Between Q1 2025 and Q1 2026, LPDDR4 memory prices rose 250%. LPDDR5 climbed 220%. In some markets, DDR5 jumped more than 400% in a single year. Memory now accounts for over 30% of manufacturing costs in some smartphone models. For budget Android devices, that share has climbed from around 15% to as high as 50%.

Counterpoint reported that average wholesale smartphone prices increased 14% in the first quarter while shipments declined 3.1% year over year. The pricing pressure will only intensify as inventories built before the supply crunch are exhausted. Some analysts expect smartphone models priced below $150 to disappear from the market entirely. The global share of sub-$99 handsets has already cratered to 12%, down from 18 to 19% in early 2025.

As Wang Yang, a principal analyst at Counterpoint, put it, the question for low and mid tier makers is "no longer how to grow shipments or market share, but whether to remain in the market at all."

Bifurcation Across the Vendor Landscape

The effects are not uniform. Apple and Samsung are weathering the crisis far better than vendors concentrated in budget devices. Apple locked in supply agreements early and is expected to hold 2026 shipments roughly flat, with a 5% rebound projected for 2027. Samsung's chip division is benefiting directly from the same AI memory boom that is hurting handset makers elsewhere, and its smartphone shipments are projected to decline only 4% this year, well outpacing the broader market.

The pain falls hardest on Chinese brands dependent on the sub-$200 segment. Counterpoint forecasts Transsion shipments will drop 32% this year. Xiaomi is projected to decline 28% after a 19% first quarter slide. Oppo has reportedly slashed targets by more than 20%, and Vivo by nearly 15%. In India, the sub-$100 smartphone market collapsed 59% year over year in Q1.

Regionally, IDC expects the steepest declines in the Middle East and Africa (down 20.6%) and in Asia Pacific excluding Japan and China (down 13.1%). North America holds up comparatively well with only a 6.3% decline, reflecting its concentration in the premium segment.

The 2011 Analog and What It Misses

Some observers are reaching for the 2011 Thailand flood analog, when hard disk drive production collapsed and rippled through PC supply chains for quarters. But the current shortage differs in important ways. In 2011, capacity came back online within a year because the disruption was geographically confined. Today's memory shortage is not a physical disruption. It reflects a durable reallocation of manufacturing priorities toward higher margin, AI related products.

IDC expects supply pressure to persist through 2026 and into mid 2027, with prices unlikely to return to 2025 levels within its forecast horizon. No real relief is expected until 2028, when new chip production capacity comes online and memory costs begin to normalize.

That timeline matters. It means 2026 and 2027 will likely see continued consolidation among smaller Android vendors who cannot absorb the cost increases or secure supply on workable terms. The smartphone market that emerges in 2028 may look quite different, with fewer players and a permanently higher average selling price.

Sector Implications and Credit Signals

The semiconductor sector is not monolithic here. Memory producers like Samsung's chip division and SK Hynix are enjoying a pricing windfall. Samsung's semiconductor workers recently ratified a profit sharing agreement drawn from a $26.6 billion pool. The shortage that is hurting handset buyers is minting fortunes for chip fabricators.

Smartphone component suppliers and contract manufacturers face a different picture. Companies heavily exposed to volume production in the low and mid tiers are navigating prolonged revenue pressure. Budget focused brands like Transsion, Xiaomi, and Oppo are likely candidates for earnings revisions downward over the next two to three quarters.

Credit spreads in the broader consumer electronics space will be worth watching. If you start to see HY spreads widen for handset exposed names while IG memory producers tighten, that is the market pricing in the divergence. So far, equities have largely ignored the smartphone data in favor of the AI narrative. Credit tends to be earlier and more granular in pricing distress.

The setup to watch: whether Apple and Samsung can actually gain share or merely lose less. And whether the memory shortage itself becomes a drag on AI buildouts once data center demand exhausts even the reallocated capacity. That would flip the narrative entirely.