🔍 Read the full analysis: Memory Supply Under Pressure: AI's Potential Impact On Tech Buyers on ThorstenMeyerAI.com
Get privacy and security gear 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 says memory demand is likely to exceed supply through 2030, as AI infrastructure growth competes with consumer electronics for production capacity. Memory prices have climbed sharply, while shipments of PCs and smartphones have fallen; the scale of future price and product effects remains uncertain.
SK hynix said on October 28 that demand for memory is expected to exceed supply through 2030, a forecast that could keep pressure on the chips used in PCs and smartphones as AI data centers buy more high-bandwidth memory. The South Korean chipmaker also said rising costs and shortages are expected to reduce PC and mobile-device shipments, though AI infrastructure demand should more than offset that decline for the company.
In its third-quarter results, SK hynix said supply is not keeping pace with market growth. The company cited the technical difficulty of advanced manufacturing and the long time needed to build production facilities. Its forecast of an imbalance through 2030 is the company’s own outlook, not a guarantee of how long prices will remain high. Chief Financial Officer Kim Woo-Hyun said on the earnings call that higher memory costs and supply shortages are expected to weigh on PC and mobile shipments.
SK hynix said demand for high-bandwidth memory, or HBM, from customers is more than twice what it can currently supply. It also said long-term agreements with 14 major customers, generally lasting five years, cover more than half of its production capacity for 2027 and 2028. The agreements include deposits, tying up part of future output in advance. The company reported third-quarter revenue of 104.6 trillion won and an operating margin of 80%, according to the source report.
Consumer-market indicators show pressure already. A 32-gigabyte DDR5-6000 kit in the United States was listed at a minimum of $447 on October 29, compared with a historical low of $72 recorded by Tom’s Hardware; the price index was updated October 26. TrendForce reported that contract prices for conventional memory chips rose 93% to 98% in the first quarter of 2026 and continued rising in subsequent quarters. These measures have different scopes: the kit price is a retail example, while TrendForce tracks chip contract prices.
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
Higher Memory Costs Reach Consumers
Memory is a component in nearly every computer and smartphone, so sustained supply pressure can affect both the price and specifications of devices. The source report cites IDC figures showing PC shipments fell 20.1% and smartphone shipments fell 22.0% in the third quarter. It says the smartphone decline was the steepest quarterly drop IDC had measured. IDC expects the average smartphone price to reach $581 this year, up 27.6%, though the figures do not establish that memory costs alone caused the changes.
Device makers may respond by charging more or offering less memory and storage at a given price. The report, citing TrendForce and Counterpoint, says some manufacturers have reduced storage, camera features or memory capacity in mainstream models, and some phone makers have reused recycled memory chips. Those choices can shape buyers’ options even when sticker prices do not rise as quickly as component costs.
The effect is not uniform across the industry. AI chip customers compete for production capacity, but suppliers also decide how to allocate it among products. Higher-margin HBM can make it commercially attractive to devote more wafer capacity to AI systems rather than standard memory. That can tighten supply for other buyers, but the precise effect on any device’s price depends on contracts, inventories, product design and manufacturers’ pricing decisions.
Products Worth Considering
Samsung Galaxy Tab S10 FE 128GB Wi-Fi Android Tablet, Blue
Apple iPad Air 11-inch (M2), 512GB, Wi-Fi, Space Gray
EcoFlow Smart Home Panel 2 Whole-Home Backup System Hub
As an affiliate, we earn on qualifying purchases.
Why AI Uses More Wafer Capacity
HBM is a form of DRAM, the working memory that holds data while a device is running. Unlike conventional memory modules, HBM stacks several memory dies and places them close to an AI processor, allowing fast data transfer. NAND flash serves a different purpose: it stores files and applications when a device is switched off. Both types of memory are used across consumer electronics and computing infrastructure.
HBM and conventional DRAM rely on overlapping manufacturing resources. TrendForce estimates that by the end of 2026 HBM will use about 22% of wafer starts from the three largest manufacturers while supplying about 9% of memory bits; by the end of 2027, it estimates about 30% of wafers and 13% of bits. The source article calculates from those figures that HBM requires roughly three times as much wafer capacity per bit. That is an author calculation based on TrendForce estimates, rather than a direct company measurement.
Other suppliers have also described tight conditions, but with different timelines. The source report says Samsung expects tight supply into 2028, while Micron has said it has no line of sight to a supply-demand balance. These are company assessments, and their forecasts may change as capacity, demand and product mixes shift.
“Supply is not keeping pace with the growth of the market.”
— SK hynix, in its October 28 earnings release
How Long Prices Stay Elevated
The length and severity of shortages remain uncertain. SK hynix’s through-2030 forecast is based on its view of demand and manufacturing constraints; it does not establish that every memory product will be scarce for that entire period. Samsung’s and Micron’s assessments use different language and horizons, and the supplied material does not provide comparable forecasts for each memory category.
It is also unclear how much of the recent increase in consumer-device costs is attributable to memory rather than other components, exchange rates, product changes or company pricing decisions. The source report presents shipment and average-price figures but does not isolate memory as the cause. Nor does it include a response from suppliers to Acer’s allegation that they are overstating shortages. The competing interests matter: manufacturers benefit from strong pricing, while large customers want predictable supply and costs.
Capacity and Pricing Over Coming Quarters
Buyers and manufacturers will be watching upcoming earnings reports, contract negotiations and production updates for signs that supply is catching up with demand. SK hynix’s long-term customer agreements cover a substantial share of its stated 2027 and 2028 capacity, but the company has not said that all future output is committed. New factories and process improvements could add supply, although SK hynix says building capacity takes time.
For consumers, the next practical signals will be retail prices and the memory and storage specifications offered in new PCs and phones. The available information does not say how much prices will change in any particular market or when conditions might ease. Those outcomes will depend on both supply growth and how AI infrastructure demand develops.
Key Questions
What did SK hynix forecast?
SK hynix said on October 28 that it expects memory demand to exceed supply through 2030, citing manufacturing complexity and the long time needed to build new facilities.
Why can AI demand affect memory for phones and PCs?
AI systems use high-bandwidth memory, which shares manufacturing resources with other memory products. If suppliers allocate more capacity to HBM, less may be available for conventional memory used in consumer devices.
Are PC and smartphone prices certain to rise?
No specific price increase is established for all devices. Memory costs and retail prices are under pressure, but the effect depends on supply, contracts, other component costs and manufacturers’ pricing choices.
What is known about shipments?
The source report cites IDC figures showing PC shipments fell 20.1% and smartphone shipments fell 22.0% in the third quarter. Those figures describe shipment changes; they do not show that memory shortages were the sole cause.
When might the supply situation improve?
There is no confirmed date for improvement. SK hynix forecasts an imbalance through 2030, while the report cites different outlooks from Samsung and Micron. Additional production capacity and changes in demand could alter those forecasts.
Source: ThorstenMeyerAI.com
