From Xiaomi to HP, multiple technology companies warn of "memory shortages," and Dell's conference call states, "We've never seen costs rise this quickly."

From personal computers to smartphones, the technology industry is facing a memory chip supply crisis triggered by the construction of AI infrastructure. Major tech giants such as Dell, HP, Xiaomi, and Lenovo have recently issued warnings, stating that the memory shortage will drive up manufacturing costs for nearly all electronic products, from smartphones to medical devices to automobiles.

Dell's Chief Operating Officer Jeff Clarke stated during an analyst conference call on Tuesday that the company has "never seen costs rise at the current pace." He pointed out that the cost base for all products, from high-bandwidth memory used for AI to dynamic random-access memory (DRAM) for personal computers, as well as hard drives and NAND flash memory, is increasing. Dell indicated that it will consider all options, including repricing some devices.

HP CEO Enrique Lores mentioned in a media interview that the second half of 2026 is expected to be particularly challenging, and the company will raise prices if necessary. He revealed that HP is taking "proactive measures," including bringing in more memory suppliers and reducing the amount of memory used in products. According to the company, memory accounts for 15% to 18% of the cost of a typical personal computer.

Market research firm Counterpoint Research predicted this month that memory module prices will rise by 50% by the second quarter of next year. Xiaomi has raised the prices of its flagship devices and expects the memory chip shortage to lead to further increases in mobile device prices next year. Xiaomi's management pointed out in a conference call that the current rise in memory prices is a long-cycle behavior, primarily driven by the surge in HBM demand brought about by AI, rather than the traditional cyclical fluctuations in the smartphone and laptop industries.

AI Boom Changes Capacity Allocation Pattern

The root of this shortage lies in the AI boom's reallocation of chip capacity. Memory chips are divided into processing auxiliary types and information storage types, and manufacturers are allocating more capacity to new, more complex, and higher-margin products used in AI systems, leading to shortages of more common types of memory.

The memory shortage is due to manufacturers prioritizing business with the world's largest AI chip supplier, Nvidia, headquartered in Santa Clara, California. Nvidia has consistently emphasized assembling the highest value and most advanced systems for AI data centers. Analysts warn that the memory shortage could limit automotive and electronic product production in 2026.

The business of logic chip suppliers may also be affected, as customers may avoid placing orders if they cannot obtain sufficient memory. These chips are used for data processing and are crucial for building AI systems.

Manufacturers Stockpiling and Raising Prices in Response

In the face of the supply crisis, several manufacturers have begun to take defensive measures. Lenovo's Chief Financial Officer Winston Cheng stated that Lenovo's memory inventory is about 50% higher than usual. ASUS has also joined the stockpiling ranks, and both PC manufacturers plan to maintain stable prices during the holiday quarter and reassess market conditions in the new year.

Xiaomi's management revealed that the company has made early arrangements and signed supply agreements with partners for the entire year of 2026 to ensure that supply remains unaffected throughout the year. Lu Weibing admitted that the company might partially absorb the pressure of rising memory costs through product price increases, while the fundamental solution lies in continuously optimizing product structure and steadfastly promoting high-end development The company's target for mobile phone shipments in 2025 remains at approximately 170 million units, but it focuses more on improving product structure rather than simply increasing quantity.

Lenovo emphasized its economies of scale and stated that it sees opportunities to capture more market share while competitors struggle by leveraging supply chain advantages. However, Cheng echoed Dell's statement, calling this cost surge "unprecedented."

Apple provided a relatively optimistic assessment. Chief Financial Officer Kevan Parekh acknowledged during an analyst conference call that there is "slight upward pressure" on memory prices, and the cost structure of some new products is slightly higher, but he emphasized that Apple is managing costs well. Similar to Lenovo, Apple, as one of the largest customers of many manufacturers in the electronics supply chain, helps it secure favorable conditions for continuous supply.

Supplier Valuations Surge Reflect Market Tension

Supply tightness has driven up the valuations of the world's top memory manufacturers. As inventory decreases and supply challenges become increasingly apparent, the stock prices of South Korea's Samsung Electronics, SK Hynix, and Micron Technology have surged in recent months. SK Hynix stated last month that it has sold out its entire memory chip product line for next year, and Micron expects supply to remain tight until 2026.

Japan's Kioxia Holdings, which focuses on NAND production, has seen its stock price multiply after going public last December, driven by the same supply tightness. Sanjeev Rana, head of research at CLSA Korea Securities, stated after Samsung's earnings report last month: "Demand for all memory-related products—whether advanced or traditional—is very strong, and supply is lagging. We may see price increases for DRAM and NAND continue for several quarters."

SK Hynix's parent company, SK Group Chairman Choi Tae-won, also warned earlier this month about supply bottlenecks. "We have entered an era where supply is facing bottlenecks," Choi Tae-won said during a keynote speech at the SK AI Summit in Seoul. "We are receiving many requests for memory chip supplies from companies, and we are working hard to think about how to meet all the demands."

The race to build AI infrastructure has already driven up energy bills in areas near large data centers. Now, this AI-driven battle for memory chips is transmitting cost pressures throughout the entire consumer electronics supply chain