Auto-Grade Memory Chip Surge Drives EV Price Hikes as AI Demand Squeezes Automotive Capacity

Over the past three months, prices of automotive-grade memory chips have surged by approximately 180%, with the increase already passed on to end consumers. More than ten Chinese new energy vehicle (NEV) manufacturers have raised prices or reduced discounts, typically by RMB 2,000 to 6,000. The primary driver behind this price hike is the strong demand for memory capacity from artificial intelligence (AI) applications. The world’s top three memory makers—Samsung, SK Hynix, and Micron—have allocated over 80% of their advanced-node production capacity to AI servers, significantly squeezing supply for automotive-grade memory chips. As vehicle intelligence accelerates, memory requirements in cars are soaring: L2-level ADAS-equipped vehicles need about 8GB of memory, while models featuring urban navigation-on-autopilot (NOA) and on-device large language models require more than 300GB—a more than 35-fold increase. Automotive-grade memory chips must meet stringent reliability and safety standards, with certification cycles lasting 18 to 24 months, causing production capacity expansion to lag behind surging demand. Rising component costs have forced NEVs to raise prices. Meanwhile, internal combustion engine (ICE) vehicles have intensified promotional efforts, yet sales have not shown significant recovery—April retail sales totaled approximately 530,000 units, down 37% year-over-year. TrendForce forecasts global NEV sales will reach 23.35 million units in 2026, up 14% year-over-year, continuing to outpace ICE vehicle growth. However, upstream cost pressures are increasingly dampening end-market demand, and price increases in the automotive market in 2025 are expected to exceed previous forecasts.