🔍 Read the full analysis: Why AI Demand Could Keep Memory Supplies Tight Through 2030 on ThorstenMeyerAI.com
Get hardware and tech essentials 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 could exceed supply through 2030, citing AI demand and the time and technical difficulty involved in expanding production. Samsung and Micron have also described a prolonged supply squeeze, but the duration and effects on consumer prices remain uncertain.
SK hynix expects memory demand to exceed supply through 2030, the South Korean chipmaker said on October 28, citing the time needed to build factories and the difficulty of producing advanced memory. The outlook matters beyond data centers: AI chips use high-bandwidth memory made from production capacity that also serves computers and phones, putting pressure on prices and availability for consumers.
In its third-quarter release, SK hynix said supply was not keeping pace with market growth and forecast that the imbalance would persist “through 2030.” The company attributed that outlook to the long lead time for new production facilities and the technical difficulty of leading-edge processes. The forecast is the company’s assessment, not a guarantee of how prices or supply will develop.
On the earnings call, Chief Financial Officer Kim Woo-Hyun said rising memory costs and shortages were expected to reduce PC and mobile-device shipments, while growth in AI infrastructure demand would more than offset that decline. An SK hynix executive also said customers’ requirements for high-bandwidth memory exceeded twice the amount the company could currently supply. Long-term agreements with 14 major customers cover more than half of its production capacity for 2027 and 2028, the company said.
Another supplier, Samsung, has said tight supply could continue into 2028. Micron has said it has no “line of sight” to a supply-demand balance. Those outlooks point to a broad industry concern, although each company’s forecast reflects its own expectations and business interests.
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 Competes With Consumer Chips
High-bandwidth memory, or HBM, is a form of DRAM stacked in layers and placed beside processors used to run AI workloads. It moves data rapidly, but its production draws on the same broad manufacturing resources used for other memory. That creates a trade-off: suppliers can direct capacity toward HBM for data centers rather than conventional DRAM used in PCs and phones.
TrendForce estimates that by the end of 2026 HBM will use about 22% of wafer starts at the three largest memory makers while providing about 9% of their memory bits. For the end of 2027, it estimates about 30% of wafer starts and 13% of bits. These figures describe the market estimates for those periods; they do not establish how much any individual product’s price will rise.
The effects are already being reported in consumer markets. The source reports that a 32-gigabyte DDR5-6000 kit cost at least $447 in the United States on October 29, compared with a lowest recorded price of $72 for that type of kit. TrendForce data cited in the source show ordinary memory contract prices rose 93% to 98% in the first quarter of 2026. Prices and availability vary by product and retailer, and the cited figures do not by themselves explain every change in device costs.
Products Worth Considering
Apple MacBook Pro 14.2 with M5 Pro Chip, 18-Core CPU / 20-Core GPU, 64GB RAM, 1TB SSD, Silver
Apple MacBook Pro 16-inch (2021) with M1 Max, 64 GB RAM, 2 TB SSD – Space Grey
Apple MacBook Pro 16.2-inch with M5 Pro Chip (18-Core CPU, 20-Core GPU), 48GB RAM, 1TB SSD, Nano-Texture Display, Space Black
As an affiliate, we earn on qualifying purchases.
How AI Uses Memory Capacity
DRAM temporarily holds data a device is actively using; NAND flash stores files and apps after a device is switched off. PCs commonly use DDR5 DRAM, while phones use memory chips soldered into the device. HBM is also DRAM, but its stacked design provides high data-transfer capacity for AI processors.
Building and equipping new chip factories takes time, while advanced processes can be difficult to scale. SK hynix says those constraints underpin its forecast. Its view is not the only one: Samsung has described tight supply into 2028, while Micron has said it cannot see when supply and demand will balance.
Demand forecasts should be distinguished from confirmed production outcomes. SK hynix said long-term contracts cover more than half its capacity for 2027 and 2028, but that does not mean every memory product is already allocated or that current shortages will continue at the same level. The company also reported record third-quarter results, giving it a financial interest in a market where high demand and constrained supply support strong returns.
““The structure in which demand exceeds supply [is expected] to persist through 2030.””
— SK hynix, in its October 28 earnings release
Forecasts Face Market Pushback
The 2030 outlook is a forecast, not a confirmed shortage timeline. Actual supply will depend on factory construction, production yields, demand from AI companies and other buyers, and changes in technology. The source does not provide a firm estimate of how much prices for specific consumer devices will rise or how long current retail prices will last.
There is also a dispute over how suppliers describe the market. Acer has accused memory suppliers of talking up shortages to protect margins, according to the source. Suppliers, meanwhile, report strong demand and constraints on capacity. The material provided does not independently resolve that disagreement. Shipment and price figures cited for PCs and phones also cover specific periods and should not be treated as proof that every market or product is experiencing the same conditions.
Watch Supply and Device Costs
Further earnings reports and company guidance should show whether memory makers maintain their long-range forecasts and how quickly new capacity comes online. Buyers and analysts will also be watching HBM demand, production commitments for 2027 and 2028, and prices for conventional DRAM and NAND.
For consumers, upcoming PC and phone launches may reveal whether manufacturers pass on higher component costs, reduce memory or storage in lower-priced models, or use other cost-saving measures. The pace and scale of those changes are not yet clear. Memory markets are volatile, and the outlook carries risks for suppliers, device makers and customers; this report is not investment advice.
Key Questions
What did SK hynix forecast?
SK hynix said on October 28 that it expects memory demand to exceed supply through 2030, citing factory lead times and the technical difficulty of advanced production.
Why can AI demand affect PC and phone memory?
AI processors use HBM, a type of DRAM that draws on manufacturing capacity also used for other memory products. Suppliers’ choices about how to allocate capacity can affect the supply of conventional memory used in computers and phones.
Have other memory makers made similar forecasts?
Yes. Samsung has said tight supply could continue into 2028. Micron has said it has no “line of sight” to a balance between supply and demand. These are company assessments, not guarantees.
Does the forecast mean PC and phone prices will keep rising through 2030?
No. SK hynix’s forecast concerns the balance between memory demand and supply. It does not specify how prices for particular devices will change, and future costs will depend on supply, demand and manufacturers’ decisions.
What remains disputed?
Acer has accused suppliers of talking up shortages to protect margins, while suppliers cite strong demand and limited production capacity. The available information does not settle that disagreement or confirm how long any specific shortage will last.
Source: ThorstenMeyerAI.com
