Rising component costs are putting pressure on both device prices and specifications. The squeeze can mean paying more, getting less memory or keeping existing hardware for longer.
A laptop can become worse value without its price moving. Keep the sticker the same, trim the memory or storage, and the buyer still loses something.
That is the consumer problem emerging from the latest memory-market research. TrendForce reports that cost pressure is pushing buyers towards lower-capacity memory modules as suppliers prioritise servers. AI infrastructure demand is helping reshape the choices available to people shopping for ordinary devices.
Our earlier coverage examined how the AI memory boom was feeding into gaming-hardware prices. The next question is what manufacturers change when customers cannot simply pay more.
The notebook number needs its conditions
In its notebook research published on 4 September 2026, TrendForce uses a mainstream laptop with a US$900 MSRP in Q1 2025 as its benchmark. It estimates that the CPU, DRAM and SSD together rose from roughly 45% of its bill of materials to 68% by Q3 2026.
Those are shares of component costs, not shares of the shop price.
TrendForce’s model estimates that a comparable product would need to cost about 80% more to preserve the same gross margin as in Q1 2025. This is a conditional modelling exercise, not a forecast that laptops will become 80% more expensive.
The distinction matters because keeping margins unchanged is a choice. A manufacturer can accept less profit, change the configuration or risk losing customers with a higher price. None of those choices makes the underlying cost pressure disappear.
Less memory is already part of the response
The public summary of TrendForce’s Q3 DRAM analysis says suppliers are prioritising server applications over PCs and smartphones, while high costs are driving buyers towards lower-capacity modules. This is industry research describing procurement behaviour, rather than a list of confirmed specification cuts across every laptop or phone range.
Even so, it points to a more awkward outcome than a straightforward price rise: a familiar budget may buy a less generous configuration.
Consider an illustrative choice between two otherwise comparable laptops, one with 16GB of RAM and a 512GB SSD, the other with 32GB and 1TB. Keeping the cheaper configuration on sale can preserve an accessible starting price. It does not preserve the same headroom for demanding software or the same room for a growing game library.
That example is not a reported product downgrade. It shows why the specification sheet matters as much as the advertised price. For anyone who needs the extra capacity, the relevant comparison is the cost of the configuration they will actually use.
The same logic applies to compromises elsewhere. A faster processor is not automatically a better upgrade if getting it means accepting too little storage or cutting another feature that matters more to you.
AI is part of the pressure, not the whole bill
TrendForce’s July memory outlook describes production capacity moving towards server applications and enterprise SSDs. The connection reaches beyond the specialist high-bandwidth memory attached to AI accelerators: conventional system memory and storage also serve the expanding datacentre market.
But that outlook also identifies production cuts and suppliers withdrawing from older consumer-memory segments. Those supply decisions matter alongside AI demand.
Other components have their own cost pressures. TrendForce’s earlier notebook research identified higher copper costs and more complex circuit boards, battery-material costs, and additional expense associated with newer connectivity standards. A notebook’s bill cannot be reduced to memory alone.
There is another useful distinction in TrendForce’s latest DRAM market bulletin: it reports stronger contract pricing alongside subdued spot trading. Prices negotiated by large buyers, transactions in the spot market and the amount a retailer charges for a finished PC are different things. One is not a live price list for the others.
The squeeze can change when people buy
TrendForce’s notebook research says inventories bought at lower costs can temporarily soften the impact on prices and margins. As those stocks run down, manufacturers face a harder choice about how much of today’s costs to pass on.
Timing is also showing up in phones. The public summary of its latest smartphone production report attributes some stronger near-term activity to consumers buying ahead of anticipated price increases. It warns that longer replacement cycles and the used-device market could become more attractive as affordability worsens.
Buying early and keeping a device longer can therefore be two responses to the same pressure. Someone who already needs a replacement may bring the purchase forward; someone whose current hardware still works may postpone it.
Neither response makes a blanket “buy now” instruction sensible. Spending money months before you need to can be an expensive way to avoid a price increase that may not reach your chosen product.
Compare the complete device, not the starting price
For buyers, the useful response is to set the minimum specification before comparing deals. Check RAM, usable storage options, the exact processor, and whether the capacity you need can be added later. A low entry price is less helpful if the configuration that meets your needs costs substantially more.
If the current machine remains adequate, keeping it is a legitimate choice. If it needs replacing, judge the available device on its complete price and useful lifespan rather than buying extra capacity solely out of fear.
The sharper test for the next hardware refresh is simple: does the new model improve what you can do, or has the same budget quietly started buying less?
