# What Exactly Did Jensen Huang Say About Nvidia's Growth?

According to TechCrunch, Jensen Huang said in a September 10, 2026 interview that Nvidia will grow "an astounding 70%" next year. Huang also insisted that Nvidia's deals with customers and partners are not circular — meaning he denies that Nvidia is effectively funding its own revenue by investing in companies that then buy its GPUs. The 70% figure is extraordinary at Nvidia's scale. Nvidia reported $130.5 billion in revenue for fiscal 2026, according to the company's fourth-quarter earnings release. A 70% increase would put Nvidia at roughly $222 billion in fiscal 2027 — a number larger than the GDP of most countries. TechCrunch reported that Huang attributed this growth to demand across every segment Nvidia serves: hyperscale data centers, enterprise AI, sovereign AI, and robotics. But the circularity question is not new. Nvidia has invested in OpenAI, CoreWeave, and other companies that are also among its largest customers. When a company invests in a customer, and that customer uses the investment to buy the investor's product, the revenue is real but the demand signal is distorted. Huang's denial is a direct response to that criticism, but the structure of the deals has not changed.

# Is Nvidia's Growth Actually Circular?

Huang told TechCrunch the deals are not circular. But the evidence is more complicated than a binary yes or no. Nvidia's investment in OpenAI — reported by multiple outlets including The Information and Bloomberg — involved Nvidia providing capital that OpenAI then used, at least in part, to purchase Nvidia GPUs. CoreWeave, a cloud provider that is one of Nvidia's largest customers, received a $2.3 billion investment from Nvidia in 2024, according to CoreWeave's SEC filings. CoreWeave's business model is built almost entirely on renting Nvidia GPUs. According to TechCrunch, Huang argued that these investments are strategic bets on the AI ecosystem, not financing arrangements designed to boost Nvidia's own sales. He said the demand for Nvidia's products is driven by end-user applications, not by Nvidia's balance sheet. That may be true in the long run. But in the short run, the distinction between "strategic investment" and "vendor financing" is blurry. If Nvidia invests $1 billion in a startup, and that startup spends $800 million on Nvidia GPUs, Nvidia's revenue increases by $800 million — but $1 billion of Nvidia's cash is now tied up in an illiquid asset. The revenue is real, but the cash flow is not as clean as it appears.

# Who Are Nvidia's Biggest Customers and How Much Do They Depend on Nvidia?

Nvidia's largest customers are Microsoft, Meta, Google, Amazon, and Oracle — the hyperscalers building AI data centers. According to Nvidia's fiscal 2026 earnings release, four customers accounted for more than 40% of Nvidia's revenue. That concentration is a risk: if any one of them slows spending, Nvidia's growth projection collapses. But the newer customers are more troubling. CoreWeave, Lambda Labs, and other GPU cloud providers are smaller, less diversified, and more dependent on Nvidia. CoreWeave's SEC filings show that Nvidia GPUs represent the overwhelming majority of its capital expenditures. If Nvidia raises prices or changes allocation, CoreWeave has no alternative. Huang's argument is that these companies have real customers — AI labs, enterprises, governments — who are paying for compute. That is true. But those end customers are themselves often funded by venture capital or, in some cases, by Nvidia's own investment partners. The chain of dependency is long, and at each link, the demand signal gets weaker.

# What Does the Comparison Table Show About Nvidia vs. Its Customers?

Company	Nvidia Investment	Nvidia GPU Dependency	Revenue Source	Risk Level
OpenAI	Yes (reported)	High	API, subscriptions, Microsoft	Medium
CoreWeave	$2.3B (SEC filing)	Extreme	GPU cloud rental	High
Microsoft	No	High	Azure, Office, OpenAI	Low
Meta	No	High	Advertising	Low
Lambda Labs	Yes (reported)	Extreme	GPU cloud rental	High
Verdict	Nvidia's growth is real, but the circularity risk is concentrated in smaller customers who cannot survive a demand pause.

# What Happens If the 70% Growth Projection Is Wrong?

If Nvidia misses its 70% growth target, the consequences would cascade through the entire AI ecosystem. Nvidia's stock price — which has driven a significant portion of the S&P 500's gains over the past two years — would fall sharply. That would hit retail investors, pension funds, and index funds. More importantly, a Nvidia miss would signal that the AI buildout is slowing. Companies like CoreWeave, which have borrowed heavily to buy GPUs, would face a debt crisis. According to CoreWeave's SEC filings, the company has billions in debt secured against its GPU fleet. If GPU demand falls, the collateral value falls with it. The hyperscalers — Microsoft, Google, Amazon — would be less affected because they have diversified revenue. But they would still face pressure to justify their AI capital expenditures. If Nvidia's growth slows, the narrative that AI is an infinite growth market weakens. Huang's 70% claim is a bet that demand will remain strong for at least another year. The evidence for that bet is mixed. On one hand, Nvidia's backlog is reportedly massive, and sovereign AI deals are growing. On the other hand, the circularity concern is real, and the customer concentration is extreme.

# Predictions

1. Nvidia will report fiscal 2027 revenue between $200 billion and $225 billion, hitting or slightly missing its 70% growth target, but will guide to less than 30% growth for fiscal 2028. 2. CoreWeave will report a revenue miss or cut guidance by Q2 2027, and its stock will fall more than 40% from its 2026 peak as GPU rental prices decline. 3. The SEC will open an informal inquiry into vendor financing arrangements in the AI hardware sector by mid-2027, focusing on whether Nvidia's investments in customers were adequately disclosed as related-party transactions.

# Article Summary

• Jensen Huang's 70% growth claim is credible for fiscal 2027 but the circularity denial is not — the deal structure tells a more complicated story.
• Nvidia's customer concentration is extreme: four customers account for over 40% of revenue, and smaller customers like CoreWeave are dangerously dependent on Nvidia GPUs.
• The real risk is not Nvidia's revenue but the cash flow quality — vendor financing inflates demand signals and masks underlying weakness.
• CoreWeave is the canary in the coal mine: if GPU rental demand slows, its debt-heavy model collapses first.
• The SEC is likely to scrutinize vendor financing in AI hardware by 2027, creating a new regulatory risk for Nvidia and its partners.