🔍 Read the full analysis: AI's Memory Boom: What It Means For Devices And IT Spending on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
SK hynix forecast on Oct. 28 that memory demand will exceed supply through 2030, citing difficult production processes and long factory build times. The company says AI infrastructure demand is expected to offset lower PC and phone shipments, while rising memory costs are already pressuring consumer-device prices and specifications.
SK hynix said on Oct. 28 that memory demand is likely to exceed supply through 2030, a forecast that underscores how AI data-center construction is competing with PCs and phones for memory-chip production. The South Korean supplier also said rising prices and shortages could reduce shipments of personal computers and mobile devices.
In its third-quarter release, SK hynix cited the technical difficulty of advanced manufacturing and the long lead time for new factories as reasons the supply gap may persist. The company said supply is not keeping pace with market growth. Its 2030 outlook is a company forecast, not a guarantee that shortages or price increases will continue at a particular rate.
On the earnings call, Chief Financial Officer Kim Woo-Hyun said higher memory costs and supply constraints were expected to weigh on PC and mobile-device shipments, while demand for AI infrastructure would more than offset that decline. An executive said customers’ requested volumes for high-bandwidth memory (HBM) were more than twice what SK hynix could currently supply. The company also said long-term agreements with 14 major customers cover more than half of its production capacity for 2027 and 2028.
The pressure reaches beyond specialized AI chips. HBM is made from DRAM, the same broad category of memory used for a computer’s working memory and in many devices. Industry estimates cited by the source suggest producing a given amount of HBM can use substantially more wafer capacity than ordinary memory. Manufacturers deciding how to allocate constrained production may favor HBM, which is in high demand from AI-chip customers.
Infographic 01 · Where the memory goes
AI data centers get first call on memory.
High-bandwidth memory (HBM) for AI chips and the memory in your phone come out of the same factories. HBM needs far more wafer per gigabit, and its buyers sign years ahead.
The bakery problem
HBM share of wafers vs. share of the memory bits they produce
Same factories, same wafers: every wafer used for AI memory is missing from phone and PC memory.
Who gets first call
Long contracts lock up supply years ahead
- SK hynix: deals with 14 major customers, usually five years, with deposits; more than 50% of 2027–2028 capacity
- Micron: more than 75% of fiscal 2027 shipments already committed
- HBM demand: customers want “more than double” what SK hynix can supply today
- Cloud budgets: memory ≈ 47% of big cloud providers’ capex in 2026, 68% in 2027 (TrendForce estimates)
Three makers, three horizons, one customer’s view
- SK hynixThrough 2030Demand above supply (company forecast)
- SamsungInto 2028Tight supply, based on market research
- MicronNo “line of sight”Tighter in 2027 and 2028 than in 2026
- Acer (customer)Easing from 2H 2027Says suppliers talk up the shortage
HBM is a wedding cake, phone memory is bread. Same ovens. The cake customers pay more and order years ahead.Thorsten Meyer · my analogy
Sources: TrendForce, 2 Jun (wafer and bit shares, estimates); TrendForce, 25 Aug (capex shares, estimates); SK hynix, 28 Oct 2026; SK hynix call, 28 Oct; Micron call, 30 Sep; The Register, 1 Oct; Samsung call, 29 Oct; Acer via Tom's Hardware / DigiTimes. Wafer-per-gigabit ratio: my arithmetic. All horizons are forecasts.
Powered by Thorsten Meyer AI
How AI Demand Reaches Device Prices
The shortage matters because memory is a component in phones, laptops, PCs and servers, not just AI data centers. When its price rises, manufacturers may face a choice between absorbing higher costs, raising retail prices or changing product specifications. Those effects can reach consumers even if they are not buying AI hardware.
The source cites TrendForce figures showing ordinary memory contract prices rose 93% to 98% in the first quarter of 2026. It also reports that prices continued rising in later quarters, though the full-year comparison partly relies on forecasts and calculations by the source author. Those figures describe contract pricing, not a guaranteed one-for-one increase in store prices.
Device makers have other ways to manage costs. The source reports that some companies are reducing memory or storage capacity and adjusting camera specifications in mainstream models; it also cites reports of reused memory chips in some phones. Such changes could mean consumers pay more for a device with similar specifications, or receive fewer features at a similar price.
Popular right now
Bose QuietComfort Wireless Bluetooth Headphones with Active Noise Cancellation, Mic, USB‑C Charging, Deep Bass, 24‑Hour Battery – White Smok
KitchenAid Artisan 5-Quart Tilt-Head Stand Mixer, Almond Cream
Bose QuietComfort Wireless Noise-Cancelling Over-Ear Headphones, Black
As an affiliate, we earn on qualifying purchases.
The HBM Capacity Trade-Off
DRAM temporarily holds data a device is using; NAND flash stores files and apps when power is off. HBM is a form of DRAM built by stacking memory chips and placing them close to AI processors, allowing rapid data movement. These products serve different technical needs, but production draws on overlapping manufacturing resources.
TrendForce estimates cited in the source indicate HBM will account for about 22% of wafer starts at the three largest makers by the end of 2026, while supplying about 9% of memory bits. For the end of 2027, its estimates are about 30% of wafer starts and 13% of bits. Those are market-research estimates, and the source’s claim that HBM uses roughly three times as much wafer per bit is its own arithmetic based on them.
Other suppliers have also signaled tight conditions. The source says Samsung expects supply constraints into 2028 and Micron has said it has no line of sight to a balanced market. On Oct. 29, Samsung reported a record for its memory business, while its mobile and network business lost money, according to the supplied account. The differing business results show why the same shortage can benefit chip suppliers while adding cost pressure for device makers.
“Supply is not keeping pace with the growth of the market.”
— SK hynix, in its Oct. 28 earnings release
How Long the Shortage Will Last
The duration and scale of the supply gap remain uncertain. SK hynix’s forecast through 2030 reflects its view of demand, production constraints and factory timelines. Samsung’s expectation through 2028 and Micron’s comments point to tight supply, but do not establish exactly when prices or availability will normalize.
There is also a dispute over how to interpret suppliers’ warnings. The source says Acer has accused memory suppliers of talking up the shortage to protect margins. That is an allegation by a customer, not an independently established explanation. Suppliers, meanwhile, have financial interests in a market where the source reports record margins for SK hynix and a record memory result for Samsung.
Shipment and price figures need careful reading. The source reports IDC estimates of a 20.1% drop in PC shipments and a 22.0% drop in smartphone shipments in the third quarter, along with a forecast that the average smartphone price will reach $581 in 2026, up 27.6%. The supplied material does not specify all comparison details for those shipment figures, and those numbers do not by themselves prove that memory costs caused the changes.
Signals to Watch in 2027
Upcoming company results and market updates should show whether suppliers can raise output and whether demand for AI systems continues to absorb new capacity. A key milestone is the production capacity already covered by SK hynix’s agreements for 2027 and 2028; the contracts may give customers greater certainty of supply, but they do not resolve constraints across the industry.
Readers can also watch for updated estimates from TrendForce and IDC, changes in memory contract prices, and device makers’ announced specifications and prices. Those indicators can help separate the cost of memory from other factors affecting retail devices. For now, the clearest confirmed development is the supplier’s warning that demand may exceed supply for years; the precise effects on prices, shipments and product features remain unsettled.
Key Questions
What did SK hynix forecast?
SK hynix forecast that memory demand will exceed supply through 2030, citing difficult advanced manufacturing and the time needed to build new facilities.
Why does AI demand affect phone and PC memory?
AI processors use high-bandwidth memory, which is made from DRAM and draws on overlapping manufacturing resources with memory used in consumer devices. Competition for limited production capacity can contribute to tighter supply and higher costs.
Does the forecast mean device prices will rise by the same amount as memory prices?
No. Memory is one component of a device’s cost, and retail prices also depend on other components, company pricing decisions and demand. The supplied figures for memory contract prices do not translate directly into an equivalent change in phone or PC prices.
When could memory supply improve?
No firm date is established. SK hynix expects the imbalance to persist through 2030, while Samsung has indicated tight supply into 2028. These are company outlooks, and conditions may change as production and demand develop.
Source: ThorstenMeyerAI.com
