130 per cent. That is how much the combined price of DRAM and solid-state drive (SSD) storage is projected to rise by the end of 2026, according to research published on 26 February by Gartner. In real terms, that means a computer that cost you $500 last year could cost $585 by Christmas. For buyers at the bottom of the market, the maths is even bleaker: Gartner says entry-level PCs under $500 will simply cease to exist as a product category by 2028.
The price shock is forecast to drive global PC shipments down 10.4% this year compared with 2025, the steepest annual contraction in over a decade, as consumers and businesses hold onto existing hardware rather than upgrade. Ranjit Atwal, senior director analyst at Gartner, put it plainly: "This is the steepest contraction in device shipments witnessed in over a decade. Higher prices will narrow the range of devices available, prompting buyers to hold on to devices for longer, fundamentally altering upgrade cycles."
The mechanism driving this is straightforward, even if the consequences are not. Gartner projects that memory costs will climb from 16% to 23% of a PC's total bill of materials this year, a shift large enough to eliminate vendors' ability to absorb costs on low-margin products. "This sharp increase removes vendors' ability to absorb costs, making low-margin entry-level laptops nonviable," Atwal said. "Ultimately, we expect the sub-$500 entry-level PC segment will disappear by 2028."
Follow the money and a different picture emerges. The root cause is not some mysterious market failure; it is a deliberate capital allocation decision by the world's three dominant memory manufacturers. The shortage is driven in part by a reallocation of manufacturing capacity away from consumer electronics toward high-margin memory solutions for AI. Instead of expanding conventional DRAM and NAND used in smartphones, PCs, and other consumer electronics, major memory makers have shifted production toward memory used in AI data centres, such as high-bandwidth memory and high-capacity DDR5. This is a zero-sum game: every wafer allocated to an HBM stack for an Nvidia GPU is a wafer denied to the LPDDR5X module of a mid-range smartphone or the SSD of a consumer laptop.
The three companies making these calls are Samsung, SK Hynix, and Micron. Micron has already exited the consumer memory brand it ran under the Crucial name, telling the market that "the AI-driven growth in the data centre has led to a surge in demand for memory and storage," and that it has made "the difficult decision to exit the Crucial consumer business in order to improve supply and support for our larger, strategic customers." In practical terms, this leaves Samsung and SK Hynix as the only major suppliers feeding the consumer DRAM market at scale, and as most people know, less competition almost always means higher prices.
Making matters worse, the scarcity is attracting opportunists. The Register reports that online security firm DataDome has identified a large-scale bot campaign submitting more than 10 million web-scraping requests to locate available DRAM stock across e-commerce sites. According to DataDome, the bots have been hitting DRAM product pages at nearly six times the rate of legitimate users, using a technique called cache busting to ensure they receive the most current inventory data. The goal is to identify and snap up scarce DDR5 stock for profitable resale, further depleting consumer supply and driving market prices higher. DataDome noted the operators appear to be using AI tools to circumvent anti-bot protections, a development that adds an almost ironic layer to a crisis itself caused by AI infrastructure spending.
The ripple effects reach further than PC buyers. Gartner's forecast covers smartphones too, where shipments are projected to fall 8.4% this year. Basic smartphone buyers are expected to exit the market five times faster than premium buyers in 2026, as rising costs push consumers toward refurbished or second-hand alternatives. The signal from Gartner is that premium device makers will ride this out; the signal for everyone else is to get comfortable with their current hardware.
There is a reasonable counterpoint to the alarm. Higher prices do, eventually, incentivise more supply. Relief is structurally constrained by long lead times: building or expanding a DRAM fab typically takes two to three years to reach volume production. Some analysts point to Chinese manufacturer CXMT adding capacity, and SK Hynix's new M15X fab beginning to contribute output in the second half of 2026. Others caution that this shortage could lead to a prolonged super-cycle of memory price hikes rather than a short-term spike. The honest answer is that nobody knows precisely when relief arrives, and Micron has reportedly warned that the DRAM drought could last until at least 2028.
There is also a legitimate argument that the AI investment driving this crisis represents exactly the kind of transformative capital expenditure that generates long-run productivity gains. If large language models, autonomous systems, and AI-accelerated drug discovery deliver even a fraction of their promise, the memory reallocation may look like a reasonable short-term sacrifice. But that framing is cold comfort if you are a student in regional Australia who cannot afford a laptop for school, or a small business that needs a basic workstation to keep its books.
Gartner warns that PC lifetimes will extend by 15% for business buyers and 20% for consumers by end of 2026, a trend that will raise concerns about security vulnerabilities on ageing hardware. Older machines running out-of-support software are a genuine cybersecurity risk, not merely an inconvenience. Governments and institutions that have pushed digital inclusion as a policy priority need to reckon with the fact that the market is moving in the opposite direction.
For now, the pragmatic advice is the same whether you lean left or right on technology policy: if you need a new PC, buy sooner rather than later. Memory price inflation is expected to hit more aggressively in the second quarter of 2026, as existing DRAM inventory levels deplete. After that, the entry-level market will keep shrinking until, according to Gartner's modelling, it disappears altogether. The market is not broken, but it has chosen a winner. That winner is not the budget buyer.