Learning from history, how much does a memory price increase affect the mobile phone industry?

A storm in the memory supply chain, ignited by AI demand, is sweeping through the global mobile phone industry. According to Zhuifeng Trading Desk, Bernstein's latest report shows that a price increase cycle for memory chips driven by strong AI demand has arrived: **In Q4 2025, mobile DRAM contract prices are expected to rise by 30%-40% quarter-on-quarter, and NAND prices by high single-digit percentages, potentially continuing into the first half of 2026.** Bernstein notes that historically, **rising memory prices have had significantly different impacts on smartphones of different market positions.** Mid- to low-end models such as the Redmi series are hit hardest: **Memory costs account for over 10% of ASP, and if prices rise by 40%, Xiaomi’s smartphone gross margin may drop by 2-3 percentage points.** By contrast, high-end models like the iPhone are less affected because memory costs make up only 4% of ASP. Analysts point out that memory price increases affect camera component suppliers in differentiated and delayed ways. Camera lens suppliers are less impacted, while camera module suppliers see lagging effects of 1-2 quarters; **the impact on Apple’s supply chain is noticeably less severe than on Android’s.** AI Demand Triggers a Price Surge Cycle, Continuing into Early 2026 Unlike past cycles driven purely by supply and demand, the core driver of this price hike is explosive AI demand—each AI server requires eight times more DRAM than regular servers and three times more NAND. Facing this historic opportunity, chip giants like Samsung and SK Hynix are rapidly adjusting production layouts, shifting resources from low-margin LPDDR chips for phones to the highly profitable HBM (High Bandwidth Memory) products, which can achieve profit margins of 50%-60%, far surpassing traditional memory chips at 30%. This capacity shift has directly led to continued tight supply of mobile phone memory. Data shows that by Q4 2025, mobile DRAM contract prices will rise 30%-40% quarter-on-quarter, with NAND prices showing strong single-digit increases, while DDR5 16Gb chip monthly prices have skyrocketed by 102%. Even more stressful is that chip makers pausing price quotes has further exacerbated market panic, leaving phone manufacturers caught in a dilemma: **“buy and lose money, don’t buy and run out of stock.”** Bernstein points out this supply crunch will last at least through the first half of 2026, possibly extending into 2027. Mid- to Low-End Models Bear the Brunt, High-End Models Are Relatively Safe The impact of memory price rises on the phone industry shows a clear structural difference, a recurring theme in historical cycles. The core reason is the huge variation in memory cost as a proportion of different models’ ASP (Average Selling Price): > **Based on 2025 data, memory in high-end models like the iPhone 17 Pro Max comprises just 4% of ASP, so price increases have minimal impact.** > > **High-end Android models like the Xiaomi 17 Pro see a 7% memory cost ratio, so impact is relatively minor.** > > **For mid- to low-end Android models like Redmi, memory costs top 10% of ASP, making them the hardest hit by price hikes.** Brands focused on value-for-money face especially intense pressure. The “King of African Phones” Transsion’s financial report has already sounded the alarm: In Q3 2025, revenue grew 22.6% year-on-year, yet net profit dropped 11.06%, citing supply chain cost increases as the drag. Previously, in Q3 and Q4 2024, memory price hikes led to a two-point drop in mobile business gross margin. Xiaomi also faces clear challenges: **If memory prices rise 40%, its smartphone gross margin could fall by 2-3 percentage points.** To counter cost pressure, Xiaomi plans to pass some costs to customers via price hikes but risks sales declines. More importantly, high-end transformation is seen as the key strategy; shipments of the Xiaomi 17 series grew 30% year-on-year, with 80% being high-end models like the 17 Pro/17 Pro Max. Historically, during the memory price drop cycle from H2 2022 to H1 2023, Xiaomi’s gross margin improved significantly. But in the price hike cycle from Q4 2023 onward, gross margin dropped from 16.5% in H2 2023 to “low double-digit” levels. Compared to the mid- to low-end market, the high-end market shows greater resilience. Bernstein believes Apple, thanks to its ecosystem and brand premium, can absorb cost pressure effectively. Additionally, Xiaomi’s move upmarket has also provided a buffer: 80% of the Xiaomi 17 series are mid- to high-end models, substantially reducing the business’s sensitivity to memory costs. This divergence was evident in the 2023 price hike cycle, where gross margins of Android mid- to low-end models were generally under pressure, while high-end models’ profits remained stable. Survival Strategies in a Price Surge Cycle Facing a price surge that may persist to 2027, the phone industry is forming new survival logic. Bernstein’s report shows, from both history and current practice, three main strategies are key for companies to break through: **Transition to high-end models is the most effective buffer.** As highlighted in the report, high-end models have lower memory cost ratios and larger profit margins, helping withstand price shocks. Brands like realme are extending product lines to the 4000 yuan price band, aiming to upgrade their brand and absorb cost pressure. **Supply chain management capability determines the upper limit of risk resistance.** Leading manufacturers lock in basic capacity through long-term supply agreements (LTA) while ramping up cooperation with domestic memory makers to reduce reliance on external suppliers. Nvidia’s practice of paying sky-high prepayments for HBM capacity offers a new idea for phone makers to secure supply. **Technological innovation opens up new pathways.** On one hand, brands boost storage efficiency by promoting high-performance chips like LPDDR5X, indirectly offsetting cost pressure; on the other, AI phones present new opportunities—although AI features demand higher memory, smart scheduling algorithms and data compression can reduce actual burden, extending the life of low-capacity models. Historically, cycles of rising memory prices often accompany industry reshuffles. In the 2017 and 2023 price surges, smaller brands gradually left the market due to weak resource acquisition and difficulty passing on costs, while leading companies’ market share kept growing. This round, combined with capacity reshuffling driven by AI transformation, may make the “winners stay winning” trend even more pronounced. ~~~~~~~~~~~~~~~~~~~~~~~~ Content above originates from [Zhuifeng Trading Desk](https://mp.weixin.qq.com/s/uua05g5qk-N2J7h91pyqxQ). For more detailed interpretations, including real-time analysis and frontline research, please join [**Zhuifeng Trading Desk · Annual Membership**](https://wallstreetcn.com/shop/item/1000309).  Risk Warnings and Disclaimer The market carries risk; investments require caution. This article does not constitute personal investment advice, nor does it consider individual users’ investment goals, financial status, or needs. Users should decide if any opinions, views, or conclusions in this article apply to their own situation. Invest accordingly at your own risk.