A Dramatic Shift in PC Economics
The economics of building a personal computer have fundamentally changed. HP disclosed during its latest earnings call that memory costs have roughly doubled as a share of total PC manufacturing expenses, jumping from 15 to 18 percent last quarter to approximately 35 percent for the current fiscal year. The culprit is clear: artificial intelligence infrastructure is consuming the world's memory supply at an unprecedented rate.
"We did share last quarter that memory and storage costs made up roughly 15 percent to 18 percent of our PC bill of materials, and we now currently estimate this to be roughly 35 percent for the year," said HP CFO Karen Parkhill during the company's February earnings call. She confirmed what consumers have been suspecting for months — price increases are coming, and they're coming for everyone.
The AI Memory Vacuum
The root cause is straightforward supply and demand, but the scale is staggering. Training and running large AI models requires enormous quantities of high-bandwidth memory. Nvidia's latest GPU accelerators each contain tens of gigabytes of specialized HBM (High Bandwidth Memory) chips, and the company is shipping millions of these processors to data centers worldwide. Every one of those chips represents memory capacity that isn't going into laptops, desktops, or smartphones.
The squeeze has been building throughout 2025 and into 2026. Samsung, one of the world's largest memory manufacturers, issued its own warning about potential price increases driven by AI-induced shortages. Micron, another major chipmaker, has gone further — effectively abandoning its consumer brand Crucial to redirect manufacturing capacity entirely toward the more lucrative business-to-business market, where data center operators are willing to pay premium prices for memory that powers AI workloads.
The cascade effect touches every component in a computer. GPUs have been under similar pressure, with graphics card availability and pricing reminiscent of the cryptocurrency mining shortages that plagued gamers years ago. But the memory situation is arguably more severe because RAM is a fundamental requirement for every computing device, not just those used for specialized workloads.
Industry Response: Prices Up, Options Down
HP's response strategy involves a combination of price increases and supply chain diversification. Interim CEO Bruce Broussard acknowledged the reality of the situation while expressing cautious optimism. "I believe the market will rationalize over time," he said, adding that the company is actively seeking new suppliers and expanding lower-cost sourcing options for memory components.
Framework, the modular laptop maker that has built a following among enthusiasts for its repairable, upgradeable designs, has already raised desktop prices by up to $460 specifically because of the RAM crisis. The company's transparent approach to pricing has made it something of a canary in the coal mine for the broader industry — when Framework raises prices, other manufacturers typically follow within weeks.
For consumers, the impact is being felt at every price point. Entry-level laptops that once shipped with 8GB of RAM at affordable prices now carry higher stickers. Mid-range machines with 16GB or 32GB have seen the most dramatic increases. And at the high end, workstations and gaming PCs with 64GB or more have become genuinely expensive propositions.
The AI PC Paradox
There's a peculiar irony in the current situation. The very AI capabilities that are driving up memory prices are also creating a new category of hardware that consumers are being asked to buy. HP reported that 35 percent of its PC sales now come from so-called "AI PCs" — machines with dedicated neural processing units designed to run AI workloads locally.
But the demand picture is mixed. While HP touts growing AI PC adoption, rival Dell has been more candid about consumer sentiment, with executives acknowledging that everyday users don't particularly care about on-device AI features. The disconnect raises questions about whether the AI PC push is driven by genuine consumer demand or by manufacturers seeking higher margins on premium hardware at a time when basic components are already pushing prices upward.
When Will Relief Come?
Industry analysts are divided on the timeline for normalization. Building new semiconductor fabrication capacity takes years, and the current facilities are being pulled in too many directions simultaneously. High-bandwidth memory for AI accelerators, standard DDR5 for PCs, and mobile LPDDR5X for smartphones all compete for the same underlying manufacturing processes.
Some relief may come from next-generation memory architectures that promise to deliver more capacity from less silicon, but these technologies are still years from mass production. In the near term, the market dynamics are clear: AI demand will continue to grow, memory supply will remain constrained, and PC buyers will pay the difference. The era of cheap, abundant computer memory appears to be over — at least for now.
This article is based on reporting by Engadget. Read the original article.
