🔍 Read the full analysis: AI Is Driving Memory Demand; Here's What That Could Mean For Buyers on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
SK hynix says memory demand is likely to outstrip supply through 2030, while Samsung and Micron have also warned of prolonged tightness. AI data centers are increasing demand for high-bandwidth memory made using capacity that also serves everyday electronics, contributing to higher prices and possible cuts to device specifications. The duration and size of the impact remain uncertain, and suppliers’ forecasts come as they benefit from strong memory prices.
SK hynix said October 28 it expects memory demand to exceed supply through 2030, a forecast that could mean continued pressure on prices for computers and phones as AI data centers take more of the industry’s production capacity. Samsung reported a record for its memory business the following day, while its phone and network business lost money, underscoring how unevenly the tight market is affecting electronics companies.
SK hynix, one of the world’s largest memory-chip makers, said in its third-quarter results release that supply is not keeping pace with market growth. It cited the difficulty of producing advanced chips and the time required to build new facilities. The company’s forecast is its own outlook, not a guarantee that shortages or high prices will last for a precise period.
On the earnings call, Chief Financial Officer Kim Woo-Hyun said higher memory costs and supply shortages were expected to reduce PC and mobile-device shipments, while growth in AI infrastructure demand would more than offset that decline. SK hynix also said customers’ requested volume of high-bandwidth memory, or HBM, is more than twice what it can currently supply. Long-term agreements with 14 major customers cover more than half of the company’s production capacity for 2027 and 2028, according to the call.
The source reports that a 32-gigabyte DDR5-6000 kit in the United States cost at least $447 on October 29, compared with a historical low of $72 recorded by Tom’s Hardware. That is a price example for one product, not a measure of every memory product or market. TrendForce data cited by the source show ordinary memory contract prices rose 93% to 98% in the first quarter of 2026; the source says prices continued rising in later quarters, though its estimate of the cumulative increase partly relies 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
How AI Orders Reach Retail Prices
The squeeze matters because HBM and everyday memory draw on overlapping manufacturing capacity. HBM is a form of DRAM built from stacked chips for fast data transfer beside AI processors. Regular DRAM serves computers and phones, while NAND flash stores files and apps. When manufacturers devote more wafer capacity to HBM, less may be available for other memory types, though the exact effects vary by product and supplier.
That pressure can reach buyers through higher component costs, higher device prices, or changes to product specifications. The source cites market research reports that manufacturers have reduced memory or storage in some mainstream devices and that some phone makers have reused recycled memory chips. These examples do not establish that every model has changed, or that memory alone caused any particular retail price increase.
The potential effects extend beyond individual shoppers. PC and phone makers face a choice between absorbing cost increases, passing them on, or adjusting products. Businesses replacing large fleets could also pay more or defer upgrades. The reported shipment declines—20.1% for PCs and 22.0% for smartphones in the third quarter—are attributed to IDC in the source; the figures alone do not establish that memory prices caused the declines.
Products Worth Considering
Amazon Kindle Scribe Colorsoft 64GB (1st Generation)
Apple iPad Air (2025) 13‑inch Wi‑Fi 256 GB – Space Gray (M3 chip)
Samsung 27-Inch Odyssey QD-OLED G8 4K 240Hz Gaming Monitor
As an affiliate, we earn on qualifying purchases.
Why HBM Competes With Device Memory
Memory chips serve different purposes. DRAM is temporary working memory used while a device runs; NAND flash retains data when it is turned off. HBM is a specialized form of DRAM made by stacking chips, allowing data to move quickly between memory and an AI processor. That design makes it useful in data centers training and running large AI systems.
TrendForce estimates cited in the source say HBM will use about 22% of wafer starts at the three largest suppliers by the end of 2026, while accounting for about 9% of memory bits produced. For the end of 2027, its estimates are about 30% of wafers and 13% of bits. The source author calculates from those figures that HBM requires roughly three times as much wafer capacity per bit. That is the author’s calculation, not a direct quote from TrendForce.
Other suppliers have described a prolonged tight market. The source says Samsung expects constraints into 2028 and Micron has said it has no “line of sight” to a supply-demand balance. These are company assessments with different wording and time horizons, not a shared industry commitment or independently established timeline.
““The structure in which demand exceeds supply” is expected “to persist through 2030.””
— SK hynix, in its October 28 earnings release
The Forecasts Face Conflicting Incentives
The outlook is not settled. SK hynix, Samsung and Micron are suppliers whose earnings benefit when memory prices and margins are high. Their forecasts are relevant evidence about orders and capacity, but they are not neutral guarantees of future shortages. The source also reports that Acer has accused suppliers of talking up scarcity to protect margins; that is an allegation, and the material does not provide evidence resolving the dispute.
It is also unclear how quickly new production facilities will come online, how much capacity will be allocated to HBM, or whether AI infrastructure demand will match suppliers’ expectations. The cited prices and shipment figures cover particular products, regions or reporting periods. They should not be read as a precise prediction of what any individual buyer will pay or when prices might fall.
Capacity Plans and Prices to Watch
Buyers and device makers will be watching upcoming company results for updates to capacity, customer orders and memory pricing. SK hynix’s multi-year customer agreements indicate that a substantial share of its output for 2027 and 2028 is already committed, but the source does not specify how much memory will reach consumer-device makers or retail markets.
For consumers, the next useful evidence will be actual prices and specifications across new PC and phone releases, rather than forecasts alone. Companies may respond to higher component costs with price increases, lower memory or storage configurations, or other product changes. The scale and timing of those responses remain unsettled as suppliers expand production and demand evolves.
Key Questions
Why is AI demand affecting memory prices?
AI data centers use high-bandwidth memory alongside processors. HBM and conventional memory rely on overlapping wafer and manufacturing capacity, so greater HBM production can constrain supply for other uses. The size of that effect depends on production choices and total capacity.
How long does SK hynix expect the shortage to last?
SK hynix said it expects demand to exceed supply through 2030. This is the company’s forecast, not a confirmed timetable for when prices will normalize. Samsung and Micron have issued their own warnings, with different time horizons and wording.
Are PC and phone prices certain to rise?
No. Higher memory costs can put upward pressure on prices, but manufacturers may absorb costs or change specifications instead. The source gives examples of high memory prices and reported product changes, but it does not establish what every buyer or model will experience.
Should buyers wait to purchase a device?
The information cited does not support a single recommendation for all buyers. Prices and specifications vary by product, and the duration of supply tightness is uncertain. Buyers can compare current prices and configurations while treating supplier forecasts as projections rather than guarantees.
Source: ThorstenMeyerAI.com
