You’ve probably heard people say that AI is coming for your job. But right now, it’s coming for your laptop budget.
The latest tech panic has a fittingly dramatic name: RAMageddon. It refers to the global memory-chip supply shortage pushing up the price of laptops, phones, graphics cards and other devices.
AI data centers need enormous amounts of memory, especially a premium type called high-bandwidth memory, or HBM. Big Tech demand is massive, margins are better and chipmakers are following the money. The AI industry has become its main goal, pulling supply toward servers and away from regular consumer tech.
So what does this mean for you, and how could it affect your next laptop or smartphone purchase? Let’s dive deeper.
What is RAMageddon?
RAMageddon is the nickname for the AI-driven memory shortage affecting consumer electronics.
Before we get all technical, a few memory terms might help explain what’s happening.
- RAM, or random access memory, helps your device juggle tasks. More RAM usually means smoother multitasking, faster app switching and better performance when you have too many browser tabs open because apparently we all live like that now.
- DRAM, or dynamic RAM, is the working memory used by computers, smartphones, servers and graphics cards.
- NAND is the storage memory used in smartphones, flash drives and solid-state drives, which are the storage drives inside most modern laptops and desktops.
- HBM, or high-bandwidth memory, is a faster, more expensive type of memory used in AI chips and high-performance data centers.
That last type of memory is where AI changes the market. AI accelerators in data centers rely on HBM because it provides chips with fast access to data during training and running large AI models. As AI demand rises, memory-makers have more reason to prioritize parts for data centers rather than the cheaper parts used in everyday consumer electronics.
PC builders and chip nerds aren’t the only ones who will feel RAMageddon. It can affect the price of the phone, the laptop your kid needs for school and the graphics card you were hoping to buy without spending several hundred dollars more than expected.
Why AI data centers are eating the world’s memory
The memory-chip market is dominated by a small group of companies, mainly Samsung, SK Hynix and Micron, aka the Big Three. Before the AI boom, those companies supplied huge amounts of DRAM and NAND for PCs, phones, gaming consoles and servers. AI changed the math.
Jitesh Ubrani, research manager for IDC’s Worldwide Device Trackers, tells CNET, “There is a lot of memory out there. It’s more of an issue of allocation.”
HBM is more valuable than standard consumer memory because AI companies need it for the expensive accelerators powering tools like ChatGPT, Gemini and other large models. As a result, manufacturers have more incentive to prioritize HBM and server-grade products over lower-margin chips for consumer devices.
“The profit margins on the memory that goes into data centers tend to be a lot higher,” Ubrani says. “And so, that’s why we’re seeing more memory being allocated to those companies, which doesn’t leave enough capacity for the companies that are making (consumer) devices.”
Micron’s move shows how serious that shift has become. In December, the company announced it would shutter its Crucial consumer division, the brand many PC builders know for RAM and SSD upgrades. Micron said AI-driven data center growth had increased demand for memory and storage, and that leaving the consumer business would help it support larger strategic customers.
That doesn’t mean consumer memory disappears overnight. But it does mean one of the biggest memory companies is openly choosing higher-growth data center customers over retail consumer products.
OpenAI’s sprawling Stargate AI data center in Abilene, Texas.
How AI server demand became your laptop’s problem
The AI memory crunch has been building for a while, but 2026 is when the price volatility started hitting regular buyers harder. Counterpoint Research says DRAM prices jumped by 80 to 90% in the first quarter of this year compared with the previous quarter.
The shortage is already changing server-buying habits. Some supply-chain reports say server-makers are shipping systems with empty or partially populated memory slots to keep orders moving, while customers wait to add more memory later. Some call this “ghost RAM,” and it shows how far the squeeze reaches beyond laptops and smartphones.
That server-side pressure eventually lands in the consumer aisle. As memory gets more expensive, device-makers may raise prices, trim memory or storage, delay new models or do a mix of all three.
“At the end of the day, prices are going up for consumers, whether they like it or not,” Ubrani says. “It’s a combination of fewer budget options, downspecing and sometimes just having to increase the price.”
It’s almost like an AI tax. You may never even use a chatbot on your laptop, but the AI boom can still affect how much that laptop costs.
The budget PC could disappear
RAMageddon may hit budget buyers hardest. According to industry analyst Gartner, memory is expected to make up 23% of the cost of building a PC in 2026, up from 16% in 2025. That may not sound dramatic, but low-cost laptops run on thin margins. When memory eats a bigger chunk of the build cost, manufacturers have less room to offer cheap machines.
“The very low cost options of the PC market are disappearing,” Ubrani says, giving Chromebooks as an example. “Where Chromebooks used to cost $200, $300, maybe $400, but now because RAM is so expensive, vendors don’t want to build Chromebooks anymore, or they’re building fewer Chromebooks.”
Gartner also forecast that the sub-$500 entry-level PC segment will disappear by 2028. The laptop that used to cost $399 or $499 may become harder to find, less capable or both.
Instead of refreshing entry-level lines, companies may focus on premium laptops where they can protect margins. That means fewer good, cheap options and more pressure on consumers to spend $800 or more.
Businesses may stretch their laptop replacement cycles, too. Instead of replacing employee computers every three or four years, IT departments may keep older machines longer and upgrade only the workers who need more power. That could boost demand for refurbished laptops, used parts and memory upgrades.
Check the memory before you buy
You know how your favorite bag of potato chips stayed the same price but feels half-empty? We’re seeing that same shrinkflation in laptops. To keep a computer near that $499 price, manufacturers may cut corners in places most shoppers don’t check first.
A laptop line that moved to 16GB as the base configuration could slide back to 8GB. A tablet refresh could keep the same price but drop RAM. A phone-maker could reserve higher memory configurations for premium models while making the base version less future-proof.
That doesn’t always mean the device is bad. But it means you’ll need to check how much memory it has, instead of assuming a new model is automatically better than last year’s version.
AI PCs make the problem even stranger
And here comes the ironic kicker: As companies race to add AI features to computers, they’re pushing a new class of AI PCs. But these AI features require more memory to work properly. Microsoft’s Copilot Plus PCs require at least 16GB of RAM, while Apple has also pushed newer Macs toward 16GB as the baseline for Apple Intelligence.
So, at the exact moment memory is getting more expensive, the industry is trying to sell you computers that need more of it. Rising AI PC prices could slow adoption because shoppers may hold off on upgrading.
“Even before ‘RAMageddon,’ AI PCs did not fare as well as everyone thought they would,” Ubrani says, and adds that RAMageddon made that a lot worse. “And so we are seeing right now fewer AI PCs shipping than everyone had previously expected.”
The AI label may not be the biggest draw anyway. Ubrani says many people buying these machines are doing so “because they have a good amount of RAM and storage, not because they have AI capabilities.”
Should you buy now or wait?
If your laptop or phone works well, don’t panic-buy any new tech. But if you already planned to upgrade this year, waiting may not save you money. In this market, last year’s model might become a luxury you can’t afford to ignore.
Gartner calls this “memflation,” or memory price inflation, and estimates annual DRAM prices will rise by 125% in 2026, while NAND flash prices will rise by 234%.
Ubrani advises holding on to your device for as long as you can or making the most of it. “Now is the time to be considering alternatives, like maybe buying a used device or a refurbished device.”
Gartner’s report also suggests meaningful relief from high memory prices may not arrive until late 2027. IDC extends the forecast to 2028. “Prices are not going back down until at least 2028, and even then, we don’t think they’ll go down significantly. If you absolutely do need a device, the sooner you buy it, the better it is,” Ubrani says.
Even as manufacturers expand production, they still have to decide how much capacity goes to HBM, server memory and consumer DRAM, which means everyday devices may not get relief right away.
The bigger issue you’ll face is a change in the hardware you get for the price. While 8GB of memory was once enough for a basic computer, it likely won’t hold up as well for newer software and AI features. A late-2024 or early-2025 laptop with 16GB of RAM could be a better buy than a 2026 refresh with 8GB and a higher price.
For most people, 16GB should be the floor for a laptop you plan to keep for several years. For gaming, creative work or local AI features, 32GB may be better. The amount of memory you get is a much better indicator of how long your device will last than any AI branding you see on the box.
A word of advice: Don’t let the shiny “new” sticker fool you.
