🔍 Read the full analysis: AI Demand, Memory Supply And The Cost Of Future Devices on ThorstenMeyerAI.com
TL;DR
SK hynix said it expects memory demand to outstrip supply through 2030, citing the time and technical difficulty required to expand production. The company’s forecast, alongside rising memory prices and weaker PC and phone shipments reported in the source, points to continued pressure on device costs and specifications; the duration and scale of those effects remain uncertain.
SK hynix said on October 28 it expects memory demand to exceed supply through 2030, a forecast that could keep pressure on the cost and availability of memory used in PCs and phones as AI data centers buy more high-bandwidth memory. On its earnings call, the company also said rising memory costs and supply constraints are expected to weigh on PC and mobile-device shipments.
In its official third-quarter release, SK hynix said supply was not keeping pace with market growth. The company cited the technical difficulty of advanced manufacturing and the long lead time for new factories as reasons it expects demand to remain above supply through 2030. That is the company’s outlook, not a guarantee of how prices or device shipments will develop.
SK hynix’s finance chief, Kim Woo-Hyun, said on the earnings call that PC and mobile shipments were expected to fall because of rising memory costs and shortages, while growth in AI infrastructure demand was expected to more than offset that decline for the company. The company also said customers’ requested volumes for high-bandwidth memory, or HBM, were more than twice what it could supply at the time. Long-term agreements with 14 major customers cover more than half of its production capacity for 2027 and 2028, according to the source.
Consumer pricing is already under pressure in the figures cited by ThorstenMeyerAI.com. The site reported that a 32-gigabyte DDR5-6000 kit cost at least $447 in the United States, compared with a recorded low of $72 in Tom’s Hardware’s price index, updated October 26. TrendForce data cited in the report showed ordinary memory-chip contract prices rose 93% to 98% in the first quarter of 2026 and continued rising in subsequent quarters. The report’s estimate that prices were roughly four times their end-2025 level is the author’s calculation, partly based on forecasts.
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
AI Orders Press on Device Costs
The supply outlook matters because memory is a component in phones, laptops, desktop PCs and servers, as well as in AI systems. When manufacturers direct more production capacity toward HBM, a specialized memory used alongside AI processors, less capacity may be available for conventional memory. The report says the products draw on the same factories and silicon wafers, though HBM requires more wafer capacity per unit of memory output.
That competition can reach consumers through higher prices, delayed purchases or changes to device specifications. TrendForce and Counterpoint, as cited in the source, report that some manufacturers have reduced memory or storage capacities and camera specifications in mainstream devices to manage costs, while some phone makers have reused recycled memory chips. These are reported industry responses, not a claim that every new device will have reduced features.
The source also cites IDC figures showing third-quarter PC shipments down 20.1% and smartphone shipments down 22.0%, with the latter described as the steepest quarterly drop IDC had measured. IDC forecast an average smartphone price of $581 for the year, up 27.6%. Those figures indicate how affordability and demand may be affected, but the source does not establish that memory costs alone caused the shipment declines or the full price increase.
Popular right now
Bose QuietComfort Wireless Bluetooth Headphones with Active Noise Cancellation, Mic, USB‑C Charging, Deep Bass, 24‑Hour Battery – White Smok
De'Longhi Magnifica Evo Fully Automatic Espresso Machine with Manual Milk Frother
Bose QuietComfort Wireless Noise-Cancelling Over-Ear Headphones, Black
As an affiliate, we earn on qualifying purchases.
Why HBM Uses More Capacity
DRAM is working memory: it holds data a device is actively using. NAND flash is storage for files and apps. HBM is a form of DRAM made by stacking memory chips and connecting them to move data quickly near an AI processor. That design makes it useful for AI workloads, which require large amounts of fast data movement.
TrendForce estimates cited by ThorstenMeyerAI.com suggest HBM will account for about 22% of the wafer starts by the three largest memory makers at the end of 2026, while producing about 9% of memory bits. By the end of 2027, the estimates are about 30% of wafer starts and 13% of bits. The source author calculates from those figures that HBM requires roughly three times as much wafer capacity per gigabit as other memory. The calculation depends on the cited estimates and is not a direct company-reported measurement.
Supply responses take time. SK hynix pointed to factory construction lead times and the challenge of producing advanced memory. The source also reports that Samsung expects tight supply to continue into 2028 and that Micron said it had no “line of sight” to a balanced market. These are separate company assessments, not a shared formal forecast.
„“The structure in which demand exceeds supply” is expected to persist “through 2030.”“
— SK hynix, in its October 28 earnings release
How Long Prices Stay High
The 2030 outlook is SK hynix’s forecast; the source does not establish that every memory type will remain equally scarce or that consumer prices will rise at a steady rate. Demand, manufacturing yields, new capacity, product mix and customer orders could change the balance. Samsung’s expectation of tight supply into 2028 and Micron’s statement that it had no line of sight to balance indicate concern across suppliers, but do not settle the timing.
It is also unclear how much of the reported increase in device prices and shipment declines is attributable specifically to memory. The source presents industry shipment and pricing figures alongside the shortage forecast, but those outcomes can have other causes. Suppliers benefit when memory is scarce and prices are high; Acer’s allegation that suppliers are overstating shortages is disputed in the material and is not independently substantiated there.
Watch Supply and Device Results
The next evidence will come from subsequent company updates on factory output, HBM production, customer contracts and memory pricing. SK hynix’s forecast will be tested against actual supply and demand over time, while Samsung and Micron’s outlooks may change as capacity and orders develop.
For consumers and device makers, upcoming shipment data and product announcements should show whether manufacturers continue to raise prices, reduce specifications or adjust release plans. The source does not identify a specific date for when supply will ease, so the timing of any improvement remains open.
Key Questions
What did SK hynix forecast?
SK hynix said on October 28 that it expects memory demand to exceed supply through 2030, citing the difficulty of advanced manufacturing and the time needed to build factories.
Why can AI demand affect memory in phones and PCs?
AI systems use HBM, a type of DRAM made in the same factories as some conventional memory. The source’s account of TrendForce estimates says HBM uses a larger share of wafer capacity relative to the memory bits it produces, which can tighten supply for other products.
Are device prices certain to keep rising?
No. The source reports sharp increases in memory prices and an IDC forecast for the average smartphone price, but neither establishes how prices will move from here. Costs, supply, demand and manufacturers’ choices can change.
What has been reported about phone and PC shipments?
The source cites IDC figures showing PC shipments fell 20.1% and smartphone shipments fell 22.0% in the third quarter. Those are reported shipment changes; the available material does not show that memory shortages alone caused them.
Source: ThorstenMeyerAI.com
