iPhone 18 Pro production costs may increase by 40%

A supply chain analysis report from market research company TrendForce shows that the cost of components to produce the iPhone 18 Pro 256 GB version is expected to increase by 38% compared to the iPhone 17 Pro, mainly due to the continuous price increase of RAM and internal memory globally.

Specifically, memory chips, which only accounted for about 10% of the total cost of iPhone 17 Pro components last year, jumped to 34% in the third quarter of 2026 and are expected to exceed 40% in the first half of 2027. Meanwhile, in the past, A-series processors and screens were always the two most expensive components, but currently only account for 9% of each component in the total iPhone production cost.

According to TrendForce, the price of memory chips on the market has increased 5-7 times since the beginning of 2025 until now. However, Apple is said to not shift all the burden to consumers. They can only slightly increase product prices, accepting a partial reduction in profit margins to keep prices at an acceptable level, thereby protecting global consumption.

Besides, to offset costs, Apple can raise prices for older generation iPhones circulating on the market to distribute costs. Previously, expert Jeff Pu of analysis company GF Securities predicted that the iPhone 18 Pro and 18 Pro Max could be 250 to 300 USD more expensive than the current generation.

Meanwhile, TrendForce warns that the situation for Android phone manufacturers is even worse. Brands in the popular and mid-range segments are having their profit margins squeezed, forced to adjust retail prices or even discontinue some product lines to avoid losses.

The AI ​​fever is causing memory chips to run out. According to Digitimesmany of the world’s leading DRAM and HBM manufacturers have sold out of production through 2027. Samsung Electronics, Micron and SanDisk announced the end of NAND flash, while Kioxia and SK Hynix are finalizing their final contracts.

Theo Tom’s Hardwarethe cause of the prolonged crisis is the huge demand from AI data centers. Modern systems require large amounts of high bandwidth memory (HBM) to continuously feed data to the GPU. HBM has very high profit margins, so manufacturers are devoting production, investment and technical resources to this segment. This means they are reducing their focus on conventional DRAM memory, leading to a shortage in supply for computers, servers, and mobile devices.