Understanding this chart means understanding Nvidia's concerns! 🧐

Currently, Nvidia's long-term purchase commitments and guarantees have soared to $530.5 billion, while the actual revenue for the season is only $96 billion, resulting in a record difference of $434 billion.

Nvidia has historical warnings; during the period of 2021-2022, just before the consumer GPU bubble burst, their purchase commitments were also higher than actual revenue, followed by a severe destocking down cycle.

Many people overlook this data, which is why Nvidia is establishing a $500 billion AI fund aimed at countering the impact of this cycle. However, in reality, the custom chips XPU from various cloud companies and large model firms far outperform Nvidia. For example, the Jalapeno inference chip released by OpenAI this time.

Once their CapEx growth slows down, Nvidia's off-balance sheet long-term purchase obligations exceeding $500 billion may face pressure from prepaid default risks. This scenario is what the AI industry hopes to avoid most!

Although rarely mentioned, it still needs to be noted! 🤔

This article is sponsored by @binancezh, “Binance buys US stocks: global assets, zero time difference, one-click access!”

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。