Nvidia’s extraordinary growth forecast has given AI bulls a new headline number to celebrate. It has also revived the harder question underneath the rally: how much of the spending surge is being sustained by Nvidia’s own financial backing?

The spark came Wednesday, when Nvidia reported $96.2 billion in fiscal second-quarter revenue, up 106% from a year earlier and above Wall Street expectations. Data-center sales reached $89 billion as buyers continued building out AI infrastructure.¹

Then came the unusual move: Nvidia projected revenue would rise 70% in its next fiscal year, far beyond the roughly 44% growth analysts had expected. Jensen Huang argued the number reflects capacity, not the outer limit of demand: “Even though our demand is much greater than 70%, our supply allows us to confidently deliver 70%.”² He said the supply chain was running flat out, with Nvidia able to serve only about 70% of current-customer demand.³

The company’s case is that demand is becoming broader, rather than resting solely on the biggest cloud platforms. CFO Colette Kress said non-hyperscaler customers—including sovereign AI projects, neoclouds and enterprise deployments—would account for roughly half of the data-center business.⁴ That diversification bolsters the bullish view: AI compute is moving from experimental budgets toward a wider commercial market.

But the forecast landed amid scrutiny of Nvidia’s investments, guarantees and other capital support for AI customers. Kress conceded the criticism directly: “We recognize the scale of this support, and we know some will call this circular financing. We see it differently.”⁵ Nvidia says independent capital evaluates each deal and that its platform can be redeployed if a customer stumbles; skeptics see intertwined vendor finance as a vulnerability if AI returns disappoint.

The optimistic macro reading reached X, where Elon Musk reposted Epoch AI Research’s estimate that official US data misses much of Nvidia’s contribution, understating GDP growth by about 0.3 percentage points over the past year.Elon Musk (elonmusk)⁶ The split is now clear: Nvidia’s numbers suggest an AI buildout still accelerating, while its balance-sheet role ensures the argument over who ultimately bears the risk is not going away.