🔍 Read the full analysis: The Memory Shortage Behind AI Could Reach Your Shopping Basket on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- 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 exceed supply through 2030, as AI data centers compete with consumer devices for production capacity. The company and other suppliers warn of prolonged constraints, but their forecasts are not independent proof of how long shortages or price increases will last.
SK hynix said on October 28 that memory demand is expected to exceed supply through 2030, a company forecast that points to continued pressure on the chips used in phones and PCs as AI data centers buy more high-bandwidth memory. Its finance chief said rising memory costs and supply limits are expected to reduce PC and mobile-device shipments, although AI infrastructure demand is expected to offset that decline for the company.
In its third-quarter release, the South Korean chipmaker said supply was not keeping pace with market growth. It attributed the projected shortfall to the technical difficulty of advanced manufacturing and the long lead time for new factories. The company’s forecast is its own outlook, not a guarantee that shortages will last until 2030.
On the earnings call, Chief Financial Officer Kim Woo-Hyun said shipment volumes for PCs and mobile devices were expected to fall because of the cost burden from rising memory prices and supply shortages. He added that growth in AI infrastructure demand was expected to more than offset that decline. SK hynix also said customer requirements for high-bandwidth memory, or HBM, were more than twice the volume it could supply at the time.
The company said long-term agreements with 14 major customers, typically lasting five years, cover more than half its production capacity for 2027 and 2028. These agreements include deposits, according to the earnings-call account. That reserved capacity gives buyers greater certainty, but also limits how much production is available to other 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 Buyers
The potential effects extend beyond data centers because HBM and consumer memory draw on overlapping manufacturing capacity. HBM is a form of DRAM built in stacked layers for fast data transfer beside AI processors. Ordinary DRAM is used as working memory in computers and phones, while NAND flash stores files and apps. Manufacturers can shift capacity toward products with strong demand and higher returns, leaving less room for other types of memory.
The source reports that a 32-gigabyte DDR5-6000 kit cost at least $447 in the United States, compared with a recorded low of $72 for that type of kit. It also cites TrendForce estimates that contract prices for ordinary memory chips rose 93% to 98% in the first quarter of 2026. Those figures describe different measures: a retail kit price and changes in chip contract prices. They do not by themselves establish how much any particular phone or computer will cost.
For shoppers, pressure could appear as higher prices, fewer devices sold, or lower specifications at a given price. The source cites IDC forecasts of a $581 average smartphone price in 2026, up 27.6%, and reports shipment declines in the third quarter. These market figures signal strain, but they do not show that memory costs alone caused every price or shipment change.
Products Worth Considering
Dell XPS 13 9350 Laptop, 13.4 FHD+ Display, Intel Ultra 7 256V, 16GB DDR5, 1TB SSD
Apple MacBook Pro (14.2-inch, 2025) – M5 chip, 16GB Unified Memory, 1TB SSD, Space Black
Dell XPS 13 9350 AI Professional Laptop 13.4 120Hz FHD+ Display
As an affiliate, we earn on qualifying purchases.
Why HBM Competes for Wafer Capacity
HBM is made by stacking DRAM dies and connecting them with many small vertical links, allowing data to move quickly between memory and an AI processor. The source says this memory is produced using the same silicon wafers as other DRAM products. It cites TrendForce estimates that HBM will account for about 22% of wafer starts by the three largest makers at the end of 2026 while producing about 9% of memory bits; by the end of 2027, the estimates rise to about 30% of wafers and 13% of bits. These are estimates, not reported final production results.
That difference reflects the more involved production process for HBM. It gives manufacturers a reason to allocate capacity to AI-related orders, while consumer-device makers continue to need conventional DRAM and NAND. Building new facilities takes years, SK hynix said, so additional investment cannot immediately resolve a mismatch between demand and available output.
The source also reports that Samsung described tight supply extending into 2028 and that Micron said it had no “line of sight” to a balance between supply and demand. Their outlooks reinforce that suppliers see constraints ahead, but the timelines differ from SK hynix’s forecast and remain company assessments.
“Supply is not keeping pace with the growth of the market.”
— SK hynix, in its October 28 earnings release
How Long the Shortfall Will Last
The 2030 timeline is a forecast from SK hynix, not a confirmed end date for a shortage. Samsung’s reported expectation extends into 2028, while Micron said it had no clear view of when supply and demand would balance. Actual conditions will depend on production increases, customer orders and how much capacity manufacturers assign to HBM and other products.
There are also competing interests in the debate. Memory makers benefit when prices and margins are strong; the source reports record results for SK hynix’s memory business and says Samsung’s memory business set a record while its phone and network business lost money. Acer’s allegation that suppliers are overstating shortages is a claim, not an established finding. The available material does not quantify how much of any retail price increase is directly attributable to memory, or specify how device prices and specifications will change across brands.
Signals to Watch in Coming Quarters
The next evidence will come from subsequent earnings reports and production updates from SK hynix, Samsung and Micron, along with market data on memory contract prices, factory output and device shipments. Their planned capacity and the timing of new facilities will help show whether supply is catching up with orders.
For consumers, upcoming PC and phone launches can reveal how manufacturers respond: they may raise prices, keep prices steady while changing specifications, or absorb some cost increases. Those choices will vary by product and company. Until suppliers report actual output and market data show how prices move, the scale and duration of any effect on shopping bills remain unsettled.
Key Questions
What did SK hynix forecast?
SK hynix said on October 28 that it expects memory demand to exceed supply through 2030, citing difficult advanced manufacturing and the time needed to build new facilities. That is the company’s forecast, not a confirmed timetable.
Why does AI affect memory for phones and PCs?
AI processors use high-bandwidth memory, which is made from DRAM and competes for overlapping production capacity with memory used in consumer devices. Manufacturers’ allocation decisions can affect the availability and cost of other memory products.
Are phone and PC prices certain to rise?
No. The source reports rising memory prices and an IDC forecast for a higher average smartphone price in 2026, but it does not establish that every device will become more expensive or that memory costs alone explain price changes.
Do all suppliers expect shortages to last until 2030?
No. SK hynix forecast a supply-demand gap through 2030. The source says Samsung expects tight supply into 2028 and Micron reported having no clear view of when the market will balance. These are separate company outlooks.
Source: ThorstenMeyerAI.com
