Why Are Corporate Laptop Prices Rising Now?

NVIDIA doesn’t sell laptops, but demand for its AI GPUs has become one of the biggest drivers behind rising enterprise laptop costs. Every NVIDIA GB200 and Blackwell-generation GPU shipped into a hyperscaler data center carries a massive load of high-bandwidth memory (HBM), and that demand is now one of the dominant forces reshaping how much one pays for ordinary PCs.

TrendForce reports conventional DRAM contract prices rose 90–95% quarter-over-quarter in Q1 2026, with NAND flash rising 55–60% in Q1 and a further 70–75% in Q2, among the steepest increases on record. 

On Dell’s most recent earnings call, Vice Chairman and COO Jeff Clarke told analysts, “We’re repricing, it feels like, every day.” Dell is forced to adjust prices almost daily to keep up with unprecedented, ongoing inflation across raw materials, fuel, and critical tech components like CPUs and memory. 

How OEMs are Actually Responding

Lenovo Group has leaned on geographic diversification, spreading procurement across roughly 30 manufacturing plants worldwide. CEO Yang Yuanqing framed the moment bluntly on Lenovo’s fiscal Q3 earnings call, saying the competition now comes down to “who can secure more supply and achieve lower costs”, an advantage he says Lenovo’s scale across PCs, phones, and servers gives it over rivals. 

On the company’s most recent quarterly call, Yang said Lenovo had promised to sustain PC revenue momentum despite rising costs and had delivered, in his words, by shifting the mix “towards premium to improve average unit revenue”.

Dell moved first, implementing 15–20% price increases starting mid-December 2025, and has said it’s adjusting its configuration mix rather than raising every SKU uniformly. CFO David Kennedy said demand continues to outpace supply across GPUs, CPUs, traditional servers and PCs. Meanwhile, Jeff Clarke said, “The supply environment is as tight as we have ever seen, and input costs are moving higher.”

HP, Acer and Asus have announced comparable price increases across parts of their PC portfolios, which IDC characterizes as an industry-wide response rather than isolated decisions.

Without Lenovo’s manufacturing diversification or Dell’s enterprise-weighted customer base, smaller OEMs may be forced to rely more heavily on lower-memory configurations or lower-capacity storage to preserve entry-level pricing.

The Chain Reaction: From GPU Orders to PC Budgets

The mechanism is straightforward once traced end-to-end. 

Cloud providers are ordering hundreds of thousands of NVIDIA GPUs to build out AI training and inference capacity. Each generation requires substantially more memory than the last. NVIDIA’s Blackwell B200 packs roughly 192GB of HBM, 2.4x the H100’s 80GB.

To meet that demand, SK Hynix, Samsung, and Micron have shifted fabrication capacity toward HBM. The three companies that account for the vast majority of global DRAM production. This earns dramatically higher margins per wafer than the conventional DDR5 used in laptops and desktops.

Micron Technology’s own numbers illustrate the scale of that shift: on its most recent earnings call, CEO Sanjay Mehrotra told analysts the company’s entire 2026 HBM output is contracted, and that the industry expects tight conditions to persist “beyond calendar 2027”.

Elaborating on when relief might arrive, Mehrotra said supply would begin to improve only gradually “even in 2028”, a timeline specific to Micron’s own supply planning, echoed in similar terms by SK Hynix, whose 2026 HBM capacity is also fully committed. 

Samsung continues expanding HBM capacity as the industry works to meet AI-driven memory demand.

A Second Demand Shock: The Enterprise PC Refresh Cycle

Timing makes this especially painful for IT buyers. Enterprises are effectively being pushed to buy more memory per device at the exact moment memory has become dramatically more expensive.

Microsoft’s Windows 10 end-of-support deadline, with a free consumer ESU extension now pushed to October 2027, giving some breathing room, is still nudging enterprises towards eventual Windows 11 and Copilot+ PC refreshes, which require a minimum of 16GB of RAM and a dedicated NPU rated at 40+ TOPS.

HP’s CFO Karen Parkhill told analysts memory costs have risen “roughly 100% sequentially”, and now account for about 35% of a PC's bill of materials, double the share from the last quarter.

What This Means for IT Procurement

For CIOs and IT buyers planning refresh cycles into 2027 and 2028, the practical implication is that memory-driven cost increases and configuration trade-offs should now be built into refresh budgeting, alongside supplier diversification and earlier contract commitments where possible.

The downstream effect might soon be visible to consumers, as this enterprise-driven cost structure is bound to filter into retail pricing during India’s Diwali sales, China’s Singles’ Day, and the US Black Friday season, each event layering discounts on top of a materially higher 2026 price floor than a year earlier.

The broader lesson is that AI infrastructure is no longer influencing only cloud providers and semiconductor companies. As memory manufacturers devote more capacity to the components that power AI data centers, the economics of enterprise computing are changing too. 

A purchasing decision made by a hyperscaler building an AI cluster can now influence what a CIO pays for an employee’s next laptop.

>