Nvidia has announced a bold financial plan to fund AI data centers, securing up to $500 billion in commitments from major financial firms including Apollo, BlackRock, and Goldman Sachs. The initiative aims to create a secondary market for aging GPUs, ensuring sustained demand for the company’s hardware as it ages. According to the source, Nvidia has agreed to guarantee the value of its chips used as collateral in these deals, covering up to 25% of any shortfall if their value drops below expected levels. This move is intended to attract long-term institutional investors and provide a stable funding source for AI infrastructure. Source: techcrunch
The plan involves a unique risk management strategy, where Nvidia’s financial obligations will increase if demand for AI hardware weakens. This creates what financiers call 'wrong way' risk, as Nvidia’s commitments could grow alongside a decline in market demand. Despite this, the company argues that the initiative is different from past failures like Lucent Technologies, which collapsed during the dotcom bubble after lending to customers. Instead of bearing the full risk, Nvidia is encouraging other investors to take on the capital, while protecting a portion of its chips’ value in the future. Source: techcrunch
Nvidia’s CEO, Jensen Huang, has positioned the initiative as a way to ensure AI remains an 'investable infrastructure,' akin to railroads or airlines rather than short-lived assets like PCs. He argues that as AI adoption grows, the company’s hardware can be repurposed by different users, maintaining its residual value. This vision includes a broader ecosystem of used AI hardware, allowing startups and enterprises to access a range of options tailored to different AI needs. The plan also reflects the broader industry shift, as traditional funding methods have become less viable, with major companies like Oracle and Google taking on significant debt or issuing new equity. Source: techcrunch
