The Shortage Won't Break Nvidia. The Labs Might (darrenleung1.substack.com)

Nvidia is facing a challenging moment as significant shifts in the AI market have driven a steep sell-off since October 8, with CoreWeave and Nebius experiencing drops of around 30% and Nvidia approximately 10%. The issues stem from multiple factors, including discrepancies in OpenAI's revenue reporting and a newfound Chinese memory breakthrough, which has sparked concerns about the sustainability of cloud operators renting Nvidia’s GPUs. While some investors view the end of the memory shortage as a potential threat to Nvidia's pricing, analysts argue that the real risk lies in the financial health of the neoclouds dependent on these AI expenses. Despite these concerns, Nvidia's business model may remain resilient. The company reinforces its position through financial backing and guarantees for its cloud partners, allowing them to continually upgrade to new chips, which, in turn, increases Nvidia's revenue per gigawatt. With the growing demand for AI capabilities, Nvidia could navigate through market volatility as long as labs and AI operations continue to pay for compute resources. However, the sustainability of this arrangement hinges on creditworthiness, particularly of companies like OpenAI, which risks substantial losses. Thus, while the memory shortage impacts market perceptions, the more pressing concern is whether key players in the AI ecosystem can maintain their financial commitments amid rising costs.