Recently, during U.S. pre-market trading, shares of several semiconductor companies tied to artificial intelligence and data centers experienced notable declines, widely attributed to comments made by NVIDIA CEO Jensen Huang at a recent industry event. Speaking at a tech summit, Huang noted that while demand for AI computing power remains strong, some customers are beginning to optimize their hardware procurement strategies—shifting away from aggressive GPU ordering toward solutions that better balance energy efficiency and real-world application needs. Investors interpreted this as a signal that the current surge in AI infrastructure investment may be entering a temporary cooldown phase.As a result, pre-market prices for stocks including NVIDIA itself, AMD, and Broadcom all dropped. Concerns grew that if major cloud providers and tech firms scale back GPU purchases, near-term revenue forecasts could be negatively impacted. Huang also highlighted that ‘AI deployment is transitioning from the training phase to the inference phase,’ which requires significantly less computational power—potentially slowing the growth rate of high-end GPU sales.However, analysts caution that Huang’s remarks reflect a maturing, more rational phase of AI adoption rather than a rejection of its long-term potential. The market’s short-term reaction appears overblown, and AI is still expected to remain a key driver of semiconductor industry growth over the medium to long term.
近日，美股盘前交易时段，多家与人工智能和数据中心相关的芯片股出现明显下跌，市场普遍将其归因于英伟达（NVIDIA）首席执行官黄仁勋在近期公开活动中的言论。黄仁勋在一次行业峰会上表示，尽管AI算力需求持续增长，但部分客户可能已开始优化其硬件采购策略，不再盲目追加GPU订单，而是更注重能效比与实际应用场景匹配。这一表态被投资者解读为AI基础设施投资热潮或将阶段性放缓的信号。受此影响，包括英伟达自身、AMD、博通等在内的半导体公司股价在盘前交易中普遍下挫。市场担忧若大型云服务商和科技企业减少GPU采购，将直接影响未来几个季度的营收预期。此外，黄仁勋还提到“AI部署正从训练阶段转向推理阶段”，而推理对算力的需求远低于训练，这也可能意味着高端GPU的销售增速将趋于平缓。不过，分析人士指出，黄仁勋的发言更多是强调行业进入更理性发展阶段，而非否定AI长期增长潜力。短期波动反映的是市场情绪过度反应，中长期来看，AI仍是推动半导体行业发展的核心动力之一。
原创文章，作者：admin，如若转载，请注明出处：https://avine.cn/9736.html