英伟达大跌后罕见安抚市场：我司GPU比谷歌AI芯片领先一代

Core Viewpoint - Nvidia claims to be a generation ahead in the AI chip industry, asserting that its GPUs can run all AI models across various computing scenarios, offering higher performance, versatility, and interchangeability compared to Google's TPUs and other ASIC chips [1][4][3]. Market Reaction - Following reports of Nvidia's key client Meta considering a deal with Google for TPU deployment, Nvidia's stock dropped over 7%, resulting in a market cap loss of nearly $350 billion, while Alphabet's stock rose, reaching a new all-time high [4][5][7]. Competitive Landscape - Google has been diversifying its AI infrastructure investments, with increasing demand for both its TPUs and Nvidia's GPUs. Analysts estimate Nvidia holds over 90% of the AI chip market, but Google's TPU is gaining attention as a viable alternative [5][9]. - Google's latest TPU, Ironwood, released in April, is designed for AI inference workloads and can outperform GPUs in certain tasks due to its energy efficiency and lower operational costs [8][9]. Client Dynamics - Google does not sell TPUs but allows companies to rent them through Google Cloud, which may increase the TPU's presence on other cloud platforms due to agreements with companies like Anthropic [9][10]. - Despite the competition, Google remains one of Nvidia's largest customers, as flexibility in handling various workloads is crucial for its clients [10].