
Using history as a reference, how significant is the impact of memory price increases on the smartphone industry?

Mid-range and low-end models are the most affected, while high-end models are relatively safe: The memory of high-end models like the Apple iPhone 17 Pro Max accounts for only 4% of ASP, making them minimally impacted by price increases. High-end Android models such as the Xiaomi 17 Pro have a memory ratio of 7%, which is relatively small in impact. In contrast, mid-range and low-end Android models like Redmi have a memory ratio exceeding 10% of ASP, making them the hardest hit by price increases
A storm in the memory supply chain triggered by AI demand is sweeping across the global smartphone industry.
According to the Chase Wind Trading Desk, Bernstein's latest research report shows that the price increase cycle for memory chips driven by strong AI demand has arrived, with mobile DRAM contract prices expected to rise by 30%-40% quarter-on-quarter in Q4 2025, and NAND prices expected to rise by a high single-digit percentage quarter-on-quarter, with the upward trend likely to continue until the first half of 2026.
Bernstein stated that historical experience indicates that the impact of rising memory prices varies significantly across different segments of smartphones. Mid-range and low-end models, such as the Redmi series, are hit the hardest, with memory costs accounting for over 10% of ASP; if faced with a 40% price increase, Xiaomi's smartphone gross margin could decline by 2-3 percentage points. In contrast, high-end models like the iPhone are less affected, as memory costs only account for 4% of ASP.
Analysts pointed out that the impact of rising memory prices on camera component suppliers shows differentiated and lagging characteristics. Camera lens suppliers are less affected, while camera modules experience a lagging impact of 1-2 quarters, and the impact on Apple's supply chain is significantly less than that on the Android supply chain.
AI Demand Triggers Price Increase Cycle, Likely to Continue into the First Half of 2026
Unlike past price fluctuations solely driven by supply and demand cycles, the core driving force behind this price increase is the explosive growth of AI demand—the DRAM demand for each AI server is eight times that of a regular server, while NAND demand reaches three times.
In the face of this historic opportunity, chip giants like Samsung and SK Hynix are adjusting their capacity layouts, shifting resources from low-margin mobile LPDDR chips to high-margin HBM (High Bandwidth Memory) products, which can achieve gross margins of 50%-60%, far exceeding the 30% of traditional storage chips.
This capacity shift has directly led to a continued tight supply of mobile memory.
Data shows that the quarter-on-quarter increase in mobile DRAM contract prices has reached 30%-40% in Q4 2025, and NAND prices have also seen high single-digit growth, while the monthly increase for DDR5 16Gb chips has soared to 102%.
More critically, the actions of chip manufacturers to suspend quotations have further exacerbated market panic, putting smartphone manufacturers in a "buying means losses, not buying means shortages" dilemma. Bernstein pointed out that this supply tightness is expected to last at least until the first half of 2026, and may even extend into 2027.
Mid-range and Low-end Models Bear the Brunt, High-end Models Relatively Safe
The impact of rising memory prices on the smartphone industry shows significant structural differences, with this differentiation repeatedly occurring in historical cycles. The core reason lies in the varying proportions of memory costs to ASP (Average Selling Price) across different segments:
Based on 2025 data, the memory cost for high-end models like the Apple iPhone 17 Pro Max accounts for only 4% of ASP, resulting in minimal impact from price increases.
For high-end Android models like the Xiaomi 17 Pro, the memory cost accounts for 7%, resulting in relatively minor impact.
In contrast, mid-range and low-end Android models like the Redmi have memory costs accounting for over 10% of ASP, making them the hardest hit by price increases. For brands whose core competitiveness is cost performance, the impact of price increases is particularly severe.
The financial report of Transsion, known as the "King of African Mobile Phones," has already raised a red flag, with a 22.6% year-on-year revenue growth in Q3 2025, while net profit fell by 11.06%, directly stating that it was dragged down by rising supply chain costs. Previously, in the third and fourth quarters of 2024, due to rising memory prices, the gross margin of its mobile phone business decreased by 2 percentage points.
Xiaomi is facing similar pressures. When memory prices rise by 40%, its smartphone gross margin may decline by 2-3 percentage points. To cope with cost pressures, Xiaomi plans to pass on some costs through price increases, but it may face sales risks. More importantly, the transformation to high-end products has become a key strategy to withstand pressure, with shipments of the Xiaomi 17 series increasing by 30% year-on-year, of which 80% are high-end models such as the 17 Pro/17 Pro Max.
Historically, during the memory price decline cycle from the second half of 2022 to the first half of 2023, Xiaomi's gross margin significantly improved. However, during the price increase cycle from Q4 2023 to 2024, the gross margin fell from 16.5% in the second half of 2023 to a "low double-digit" level.
Compared to the mid-to-low-end market, the high-end market shows stronger risk resistance. Bernstein believes that Apple, with its ecological advantages and brand premium, can effectively absorb the pressure of rising memory costs.
In addition, Xiaomi has also gained buffer space through its high-end transformation, with mid-to-high-end models in the Xiaomi 17 series accounting for 80%, significantly reducing the overall business's sensitivity to memory costs. This differentiation trend has already shown signs during the price increase cycle in 2023, when the gross margins of mid-to-low-end Android models were generally under pressure, while the profits of high-end models remained basically stable.
Survival Rules in the Price Increase Cycle
In the face of a price increase cycle that may continue until 2027, the mobile phone industry is forming a new survival logic. Bernstein's report shows that, based on historical experience and current practices, three key strategies have become crucial for companies to break through:
High-end transformation is the most effective buffering measure. As emphasized in the research report, the memory cost proportion of high-end models is low, and the profit margin is large, which can significantly withstand the impact of price increases. Brands like realme have extended their product lines to the 4000 yuan price range, attempting to absorb cost pressures through brand upgrades.
Supply chain management capabilities determine the upper limit of risk resistance. Leading manufacturers lock in basic production capacity through long-term supply agreements (LTA) while increasing cooperation with domestic storage manufacturers to reduce external dependence. NVIDIA's practice of paying exorbitant advance payments to secure HBM capacity also provides mobile manufacturers with new ideas for capacity assurance.
Technological innovation opens up new paths. On one hand, manufacturers improve storage efficiency by promoting high-performance chips like LPDDR5X, indirectly offsetting cost pressures; on the other hand, the development of AI phones brings new opportunities — although AI functions require higher storage, intelligent scheduling algorithms can reduce actual load through data compression technology, extending the lifespan of low-capacity models.
From historical patterns, memory price increase cycles are often accompanied by industry reshuffling. During the two price surges in 2017 and 2023, small and medium-sized brands gradually exited the market due to weak resource acquisition capabilities and difficulties in cost transfer, while leading companies continued to increase their market share The capacity restructuring brought about by the AI transformation, combined with this, the trend of "the strong getting stronger" in the industry may become even more pronounced

