🔍 Read the full analysis: How AI Demand Could Influence The Next Wave Of Tech Spending on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
SK hynix said it expects memory demand to exceed supply through 2030, while Samsung has forecast tight supply into 2028 and Micron says it cannot see when the market will balance. AI data centers are increasing demand for high-bandwidth memory made using capacity that also serves ordinary devices, putting pressure on prices and potentially affecting PC and phone costs and specifications.
SK hynix expects memory demand to exceed supply through 2030, the South Korean chipmaker said on October 28, as investment in AI data centers increases demand for high-bandwidth memory. The forecast matters beyond the server market: the company’s finance chief said higher memory costs and supply shortages could reduce PC and mobile-device shipments, while price pressure is already reaching consumer electronics.
In its third-quarter release, SK hynix said supply was not keeping pace with market growth. It attributed the gap to the technical difficulty of advanced production and the long time needed to build new facilities, and forecast that demand would exceed supply through 2030. That is the company’s outlook, not a guarantee of how prices or output will develop.
On the earnings call, Chief Financial Officer Kim Woo-Hyun said PC and mobile shipments were expected to decline because of memory costs and supply shortages, while growth in AI infrastructure demand was expected to more than offset that decline. An executive said customers’ requested volumes for high-bandwidth memory, or HBM, were more than twice what SK hynix could supply at the time.
SK hynix also said long-term agreements with 14 major customers, typically lasting five years, covered more than half of its production capacity for 2027 and 2028. The company said these agreements include deposits. Such commitments give suppliers and customers greater visibility, but do not establish how much capacity will ultimately reach each product category or what buyers will pay.
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 Memory Demand Reaches Consumer Devices
The pressure reflects a competition for manufacturing capacity, not just a surge in demand for AI processors. HBM is a form of DRAM built in stacked layers for fast data transfer beside AI chips. It is made from silicon wafers also used to produce memory for PCs, phones and other devices. According to estimates from TrendForce cited in the source material, HBM requires substantially more wafer capacity per unit of memory than conventional products.
When manufacturers direct more capacity toward HBM, less may be available for conventional memory unless production expands or yields improve. That can feed into component prices and product decisions. The source reports that TrendForce estimates contract prices for conventional memory rose 93% to 98% in the first quarter of 2026 and continued rising in subsequent quarters. Those figures concern contract prices, not the price of every retail product.
For shoppers, possible effects include higher prices or changes to specifications, such as less storage or memory in mainstream devices. The source cites industry reporting that some makers are adjusting specifications and that some phone manufacturers are reusing recycled memory chips. The pace and extent of any changes will vary by product and company; the available material does not establish a single outcome for all buyers.
Products Worth Considering
DJI Osmo Pocket 3 Vlogging Camera with 1-Inch CMOS Sensor and 4K Video
DJI Mic 3 Wireless Microphone System (2 Transmitters, 1 Receiver, Charging Case)
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
As an affiliate, we earn on qualifying purchases.
How HBM Competes for Wafer Capacity
Ordinary DRAM is working memory used while a device is running. NAND flash stores files and apps. HBM is also DRAM, but stacks multiple chips and connects them with vertical wiring to move data quickly for AI processors. Its design and production requirements make it a higher-value product for chipmakers, while placing it in competition with other types of memory for factory resources.
TrendForce estimates cited in the source project that by the end of 2026 HBM will account for about 22% of wafers started by the three largest producers, while supplying about 9% of memory bits. For the end of 2027, its estimates are about 30% of wafers and 13% of bits. These are market-research estimates and projections, not reported final production figures.
The outlooks from memory companies differ in their time frames but point to continued tightness: SK hynix forecast a supply-demand gap through 2030, Samsung said constraints would last into 2028, and Micron said it had no line of sight to a balanced market. Samsung’s next-day results also showed contrasting performance across its businesses: its memory division reported a record while its mobile and network business lost money, according to the source material.
“Supply is not keeping pace with the growth of the market.”
— SK hynix, in its October 28 earnings release
Supply Forecasts Face Competing Interests
The forecasts come from companies that sell memory, and the source notes that tight supply has coincided with record margins at SK hynix. Acer has accused suppliers of overstating the shortage to protect margins. That is an allegation, not an established finding; the material does not provide independent evidence to resolve the dispute.
It is also unclear how quickly new factories, process improvements or shifts in demand might change the balance. The 2030 outlook is SK hynix’s forecast, while the wafer-share figures are TrendForce estimates. Neither establishes a fixed path for retail prices, device shipments or specifications. The source’s cited price and shipment figures also have different measures and time periods, so they should not be treated as a single direct measure of consumer costs.
Watch Capacity and Device Pricing
The next evidence will come from memory makers’ subsequent earnings updates and disclosures about factory investment, production capacity and customer agreements. SK hynix’s longer-term commitments for 2027 and 2028 may offer clues about how much output is already reserved, but the company’s forecast will need to be measured against actual production and demand.
For consumers and device makers, upcoming pricing and product announcements will show whether memory constraints lead to higher prices, lower specifications, delayed products or some combination. Until manufacturers report those decisions and market data establish how supply is changing, the scale and duration of the effects on PCs and phones remain uncertain.
Key Questions
What did SK hynix forecast?
SK hynix forecast on October 28 that memory demand would exceed supply through 2030, citing the difficulty of advanced production and the time needed to build new facilities.
Why does AI demand affect memory for PCs and phones?
AI processors use high-bandwidth memory, which is produced using wafer capacity that also serves other memory products. If manufacturers allocate more capacity to HBM, less may be available for conventional memory unless production expands.
Does the forecast mean phone and computer prices will rise?
Not necessarily for every device. The outlook points to potential cost pressure, and manufacturers may respond with price changes or different specifications. Actual effects depend on supply, demand and companies’ product decisions.
Are all memory suppliers forecasting shortages through 2030?
No. In the source material, SK hynix forecast a gap through 2030, Samsung said tight supply could last into 2028, and Micron said it had no line of sight to a balanced market. These are company assessments with different stated time frames.
Source: ThorstenMeyerAI.com
