According to data from market research firm IDC,Global PC shipments in the second quarter of 2026 were approximately 68.2 million units, a year-on-year decrease of 4.9%. IDC predicts that the average selling price of PCs will rise by about 20% in 2026 and may continue to climb slightly in 2027, while shipment volumes will further shrink.
The direct trigger for this market shift lies in the tight supply of memory chips. As major tech companies invest heavily in building artificial intelligence data centers, demand for high-end memory such as HBM and other DRAM products has surged dramatically.
Memory manufacturers tend to prioritize limited production capacity for lucrative AI-related customers, thereby squeezing the memory supply for ordinary PCs, smartphones, and other consumer electronics products.
Upward pressure on memory prices has ultimately been passed on to the complete system market. Computer manufacturers have been forced to raise product prices to compensate for revenue losses caused by declining sales volumes.
Jitesh Ubrani, Head of Consumer Research at IDC, pointed out thatManufacturers must raise prices to offset the impact of reduced shipment volumes in order to maintain or even expand revenue scales.
From current results, this strategy appears to be working. Although overall sales volume has declined, corporate clients and some consumers remain willing to pay for high-end models with stronger AI processing capabilities and larger memory capacities. Manufacturers are capitalizing on the so-called "AI PC" concept, using local AI computing power, dedicated acceleration units, and higher hardware specifications as justification for price hikes.
HP is a typical example:Recent financial reports show that its PC business revenue increased by 18% year-on-year, but shipments decreased by 16%. Dell and Lenovo's PC business revenues also grew by approximately 20% and 30% respectively, both demonstrating a growth pattern of "selling fewer units at higher prices."
This model currently relies primarily on support from the enterprise market. Certain industries, constrained by regulatory requirements, cannot place all their data and workloads in the cloud, necessitating the local deployment of AI-capable computers and workstations. Corporate device refresh cycles and demand for local AI computing have effectively offset consumer resistance to high-priced PCs.
However, this growth model also carries risks. Ordinary consumers are more price-sensitive and, faced with continuously rising costs for components like memory and processors, are likely to postpone upgrade plans and continue using older devices.
Furthermore, when purchasing AI PCs, enterprises must reallocate budgets among terminal upgrades, cloud services, and data center construction, adding uncertainty to future demand growth.