AI抢占产能，汽车芯片荒2.0逼近：单车成本或增加400美元，持续时间或超2021缺芯危机

Group 1 - The automotive industry is facing a more severe structural "chip shortage" crisis than in 2021, driven by the explosive growth of artificial intelligence (AI) and the resulting shift in chip manufacturing capacity towards data centers, leading to a sharp contraction in the supply of traditional storage chips relied upon by the automotive sector [1][2] - Morgan Stanley's report warns that the surge in spot prices for storage chips is quickly being transmitted to automakers, which could significantly impact profit margins in an already thin-margin industry, particularly for electric vehicles (EVs) that are heavily reliant on smart features [1][3] Group 2 - Unlike the 2021 crisis that resulted in a reduction of approximately 12 million vehicles, the current shortage is characterized by a "lower intensity but longer duration," with structural supply chain disruptions expected to last at least until the end of 2026 [2][6] - The cost increase for electric vehicles could reach up to $400 per unit, with internal combustion engine (ICE) vehicles facing an additional cost of $100 to $200, despite storage chips only accounting for about 0.5% of the total cost of goods sold (COGS) [3][6] Group 3 - The current crisis is expected to be more persistent and structurally impactful than the previous one, with a shift from shortages of analog chips or microcontrollers (MCUs) to a dominance of DRAM shortages, particularly affecting Chinese EV manufacturers due to their high reliance on storage chips for smart cockpit and driving features [6][7] - The supply bottleneck is projected to extend into 2027, with significant tightening in the availability of traditional chips as major manufacturers phase out older production processes in favor of advanced technologies needed for AI [7][11] Group 4 - Automakers are responding differently, with companies like Volkswagen and Hyundai currently reporting no significant supply risks due to prior inventory management practices, although Morgan Stanley cautions that the situation remains uncertain and could worsen unexpectedly [12][13] - The real challenge may arise post-2028, as the availability of traditional DRAM is expected to decline sharply, potentially forcing automakers to redesign vehicles that are still dependent on older technologies [14][15]