Throughout much of 2025, buying a laptop or building a PC has been, for once, a relatively wallet-friendly experience. Offers have proliferated, mid-range models have dropped in price, and overall, consumers have felt that electronics are “breathing again” after years of inflation and shortages. But beneath this apparent calm, a shift is brewing that already concerns manufacturers and analysts: memory is becoming a bottleneck once again, and 2026 could be the year when this tension leads to sustained price increases.
The reason is no mystery: Artificial Intelligence is “pulling” the supply chain at its most critical point. Training and running models require enormous amounts of memory, and the sector is prioritizing the most cost-effective and strategic types of RAM for data centers. Practically, this means some production capacity is shifting toward memories like HBM (High Bandwidth Memory), essential for AI accelerators, while consumer memory (e.g., DDR5 for PCs) becomes more exposed to volatility.
From a bargain to a scare: the memory cycle shifts phases
Memory has always been cyclical: periods of oversupply and low prices followed by adjustments, production cuts, and sharp rebounds. The difference now is the “magnet” of AI. If manufacturers can sell at higher prices and better margins to data center clients, the temptation to reserve wafers and capacity for that demand is enormous.
Recent forecasts already suggest notable increases in the short term. For example, TrendForce projects rises across various DRAM and NAND segments as supply tightens and the market rebalances. In their Q3 2025 outlook, they even consider significant hikes—up to around 45%—in certain memory types.
This doesn’t mean all RAM will rise uniformly, but it clearly indicates an environment where prices stop falling and the market tightens again.
The game changer: HBM, servers, and the domino effect
HBM memory has become the crown jewel because it powers much of modern AI infrastructure. When investment and production lines shift toward HBM, other types of memory become more “scarce.” This sets off a second wave: server and device memory could also face upward pressure.
An example cited in the market is the rise of LPDDR5X memory and its impact on shared supply chains. An analysis published in industry press mentions a scenario where AI-related demand might push up server and mobile memory prices, with aggressive hypotheses suggesting that server DDR5 could even double yearly by late 2026 amid strong demand pressure and limited supply.
Whether or not the worst-case scenario materializes, the message is that the “barometer” is already shifting: when memory enters scarcity, price hikes are unlikely to stay single-digit.
Manufacturers are already protecting themselves: stockpiling, configuration cuts, and deferred increases
In recent months, several large industry players have started discussing the issue more openly. A common tactic during these phases is proactive stockpiling: purchasing memory ahead of time to withstand a few quarters without passing all costs onto consumers. This helps keep displays stable but can also fuel the upward trend by artificially increasing demand in the short term.
International media have reported, for instance, that HP noted sharp increases in DDR5 costs over a short period and was preparing to adjust configurations and prices, with the expectation that effects on laptops would appear later—by late 2026—thanks to accumulated inventory. Lenovo is also believed to be employing similar strategies.
In plain language: there may still be “bargains” today, but part of the industry is buying time.
What could happen to laptops and consumer PCs?
If memory costs rise, there are three typical market responses:
- Raising the final price of the device.
- Keeping the price stable by cutting specifications, especially RAM and storage.
- Postponing product launches: delaying certain models or prioritizing higher-margin configurations.
The second option is what worries average consumers most because it’s a silent step back. If the goal is to keep prices psychologically stable, the ghost of an 8 GB base model returns—especially in classes where 16 GB used to be the norm. It’s not just an “enthusiast” issue: with modern Windows, multiple browser tabs, video calls, and AI-enabled tools, 8 GB can become short before many realize.
Furthermore, the impact isn’t limited to RAM. Storage (SSD) could also experience tensions: data centers absorbing capacity and industry reorienting priorities can tighten the market. Industry leaders have warned publicly about supply stresses in 2026, adding another layer to the total cost of a laptop or desktop.
The macro angle: component price hikes lead to decreased demand
When the price of a key component spikes, the market not only gets more expensive but also contracts. According to forecasts based on IDC analyses and specialized media, 2026 could see a correction in the PC market, with moderate scenarios projecting price increases of 4–6%, and more adverse scenarios suggesting rises of up to 8%, mainly driven by pressures on memory linked to AI advancements.
Such increases, combined with the end of aggressive promotional periods, could slow down purchases and extend upgrade cycles. When that happens, manufacturers tend to respond by reducing innovation in their catalogs—fewer “affordable innovations”—and increasing segmentation to protect margins.
What consumers and businesses should watch in 2026
For home users:
- If you need a new laptop in the first half of the year, it may make sense not to rush: the market could stay stable for a while, but price shifts often come in waves.
- Prioritize 16 GB of RAM as a reasonable baseline, especially if you work with many tabs, light editing, or AI tools.
- For desktops, consider upgrading before prices spike, as RAM is one of the components that responds fastest to tension.
For professional environments and sysadmins:
- Plan procurement with more margin and review upgrade budgets.
- Be aware that the cost per node could increase due to RAM and SSD prices, affecting virtualization, VDI, and workload consolidation.
- Reevaluate sizing policies: during periods of expensive RAM, optimization approaches (containers, bloat cleanup, memory limits, detailed observability) become even more critical to avoid over-provisioning.
Frequently Asked Questions
Why does AI make RAM prices for laptops and PCs go up if AI is “in the cloud”?
Because the infrastructure supporting AI consumes enormous amounts of memory (especially HBM), pushing manufacturers to prioritize data center lines and capacity, which tightens the rest of the market.
How much could RAM prices increase by in 2026?
It depends on the memory type and supply/demand scenario. Forecasts indicate significant rises across segments and ongoing pressure through 2026, with demanding server memory scenarios.
Will 8 GB laptops return as a standard configuration?
Possible in segments where manufacturers aim to keep prices low. If memory costs rise, a common response is to cut specifications to maintain entry-level prices.
Will SSDs and storage also be affected?
Yes. Demand pressures from data centers and production decisions can tighten the flash storage market, increasing the overall cost of a complete system (RAM + SSD).