Core Viewpoint - Nvidia is transforming itself into a "central bank" for the global AI industry by proactively investing in capacity and securing long-term contracts with clients, thereby shifting its business model from selling hardware to controlling the supply of AI computing power [4][9][20]. Group 1: Nvidia's New Strategy - Nvidia's new strategy involves a $100 billion investment and a plan to build 10GW of AI computing capacity, which could generate approximately $3.5 trillion in revenue over the next few years [7][16]. - The company is moving from a traditional chip sales model to a model where it actively shapes demand and locks in future orders through long-term contracts [11][18]. - This shift allows Nvidia to reduce revenue volatility and enhance its valuation logic, positioning it as a "computing power operator" rather than just a hardware seller [18][20]. Group 2: Partnership with Intel - Nvidia has made a strategic investment of $5 billion in Intel, acquiring about 4% of its shares, which will facilitate collaboration on AI-optimized solutions [22][26]. - The partnership aims to integrate Nvidia's GPUs with Intel's custom x86 CPUs, creating a comprehensive solution for AI training and deployment [27]. - This collaboration is expected to enhance Nvidia's influence in the AI computing standard and provide Intel with much-needed capital and market confidence [27][28]. Group 3: Implications for the AI Industry - The 10GW expansion and partnership with Intel are set to drive growth across the entire AI industry chain, allowing investors to diversify their exposure beyond Nvidia [29][30]. - The current wave of AI infrastructure investment is likened to historical capital expenditure cycles, indicating a long-term growth trajectory for Nvidia [31][32]. - Nvidia's strategy effectively locks in the arms race for AI computing power within its supply chain, positioning it as a key player in the industry [33]. Group 4: Risks and Challenges - Potential risks include dependency on OpenAI's financial health and demand, as well as regulatory scrutiny due to Nvidia's dual role as an investor and supplier [28][37]. - The success of the Nvidia-OpenAI partnership hinges on OpenAI's ability to achieve explosive growth, which remains uncertain given the competitive landscape [40][42]. - Nvidia's dominance in the AI chip market could be challenged by advancements from competitors like Google and Amazon, necessitating continuous innovation [37][42].

9月，英伟达宣布将向OpenAI投资1000亿美元，并签署部署至少10GW算力的意向书。 大家都知道英伟达是一家卖GPU的公司，但你可能没意识到：它正在悄悄把自己变成全球AI行业的"央行"。 10GW是什么概念？它相当于全球最大规模AI集群的数倍，意味着未来三到五年英伟达将为OpenAI提供前所未有的算力和设备支持。 这不是一次普通的供货交易，而是英伟达主动出击，用自己的资金帮助客户建设产能，再通过长期合同锁定未来订单，形成一个堪称"算力央 行"的闭环。 今天我们就来深度拆解：英伟达为什么要这样做，这对它的商业模式意味着什么，潜在的机会和风险在哪里，普通投资者又能从中捕捉到哪些 机会。 英伟达"央行式"新战略 要理解英伟达的动作，首先要回顾它过去的商业模式。英伟达长期以来是一家典型的芯片公司，设计GPU、交给台积电代工，然后卖给OEM、 云厂商或者AI公司。收入以一次性确认为主，周期性波动非常明显。 2023到2024年，AI训练需求爆发，H100、H200供不应求，英伟达的营收和利润暴涨，毛利率一度超过75%，市值突破3万亿美元，成为全球最 有价值的科技公司。 | 名称 | 代码 | 毛利率 | | --- ...