Memory chips are key components of almost every consumer-electronics device, including laptops and smartphones. You are now competing for them with the world’s largest AI data centers, and you are losing (and will continue to lose). 

Apple recently increased prices across its Mac, iPad, HomePod, AppleTV, and Vision Pro product lines, with analysts predicting increases for its flagship iPhone lineup as well. Tim Cook has described these price increases as “unavoidable,” but the pushback has been immediate. Apple earns record profits and sits on a mountain of cash. How can price increases be unavoidable? Can’t Apple just absorb the costs? Wouldn’t raising already-high prices significantly hurt demand?

These questions are natural, but they reflect a common blind spot when thinking about a company’s pricing strategy. When setting prices, Apple certainly cares about conventional demand-side factors, such as unit sales, revenues, competition, and consumers’ willingness to pay. But Apple’s prices are also impacted by supply-side factors. These factors affect how many units Apple can produce and how cost-effectively it can produce them. The current price increases are primarily supply-driven, resulting from shortage of memory chips. 

Memory chips, also called DRAM, are used to temporarily hold data in most computing devices, including laptops and smartphones. A different type of memory chip, called high-bandwidth memory (HBM), are used in large AI data centers for holding model weights and other data readily accessible to process queries to popular AI services, such as ChatGPT, Claude, or Gemini. Although DRAM and HBM differ in many ways, they draw on overlapping wafer-fabrication capacity at the same small group of manufacturers. As Micron and SKHynix note, HBM requires more wafers than conventional DRAM to produce the same memory capacity.

Recently, these manufacturers have been allocating more of their limited capacity to higher-margin HBM chips, which is leaving less and less capacity for DRAM chips. The result is a classic spillover effect: demand surges in one market are spilling over into cost surges in another market. In this case, less capacity for DRAM chips is translating into component shortages and sudden price increases with prices estimated to have roughly doubled in Q1 2026. 

An interesting implication of the spillover effect is that the end customers of consumer electronics devices are now effectively competing against giant data centers. As both types of customers compete for the same scarce capacity, data centers with higher willingness to pay are winning against the end customer. 

This dynamic also highlights how operational considerations affect pricing, something that standard pricing analysis often misses. In normal market conditions, demand is often the constraint and price becomes an instrument for firms to compete for this limited demand. Firms lower prices, offer promotions, and other incentives to attract customers and increase unit sales. When supply constraints kick in, this dynamic flips. Price becomes an instrument to bring demand in line with what firms can actually produce, in effect, allocating the scarce capacity to customers who can pay more for the same product.

None of this means that consumers should welcome higher prices. Apple could absorb higher component costs, sacrificing some of its margin. But even a company the size of Apple cannot make additional semiconductor capacity appear overnight.

Apple’s own AI strategy may be amplifying the impact of memory chip shortages. It has adopted the strategy of running AI models directly on its devices rather than exclusively in the cloud. This allows Apple to preserve user privacy but also increases memory demands on its phones. A model that runs on a phone has to fit within the phone’s memory, and more capable models generally require more of it. Its most powerful on-device model and the features it enables require devices with at least 12 gigabytes of memory, more than many existing iPhone models provide. The company is therefore being squeezed from both sides: capacity shortages make less memory available whereas Apple’s own on-device AI strategy is increasing its need for memory in premium devices.

The memory chip problem is of course not limited to Apple. Lenovo, Dell, and HP, all of which operate at considerably lower margins, have also signaled substantial price increases. IDC expects the average selling price of a personal computer to rise 18.3% in 2026, while global PC shipments fall 11.3% as the memory shortage constrains production. With the AI data center boom showing no signs of slowing down, consumers should be prepared for prolonged price increases. 

From a market analysis standpoint, the broader lesson is that we consistently underestimate spillover effects created by operational constraints. Markets that appear unrelated can become tightly connected when they rely on the same factories and production capacity. The tremendous investment in AI infrastructure is being paid for only by technology companies and their investors: the cost is beginning to appear in the price of every consumer device that needs a memory chip.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.

This story was originally featured on Fortune.com