Mark Cuban recently ignited a debate regarding the financing of the artificial intelligence boom, comparing Nvidia (NASDAQ: NVDA) to the IPO machines of the dot-com era that were “funding everyone and anyone.” In a post shared on X on July 28, the billionaire investor raised alarms about the sustainability of the current AI infrastructure gold rush.
Understanding the Dot-Com Comparison
Cuban’s commentary does not suggest that Nvidia is literally taking companies public. Instead, he draws a parallel to the internet boom of the late 90s, when IPO capital provided young startups with the runway needed to scale. Today, Nvidia is playing a similar role by pouring capital into AI model developers, cloud operators, and various other entities that underpin the broader AI ecosystem.
Inside Nvidia’s Massive Investment Portfolio
Nvidia’s financial footprint is substantial. By the end of the first quarter of fiscal 2027 (April 26, 2026), the company held $42.3 billion in private investments, with an additional $27 billion in contingent investment commitments. This portfolio is diverse, spanning major model developers like OpenAI and Anthropic, cloud operators such as CoreWeave (NASDAQ: CRWV) and Nebius Group (NASDAQ: NBIS), as well as foundational technology suppliers including Intel, Synopsys, Nokia, and Coherent.
While some of these deals provide a “double benefit”—where Nvidia sees a return on investment while simultaneously securing a customer for its hardware—the relationship is not always symbiotic. Many investments are directed toward suppliers rather than direct customers, meaning the portfolio itself is not a guaranteed engine for Nvidia’s own sales growth.
The Risk for Specialized Cloud Operators
Cuban’s warning is not a dismissal of all AI stocks; established cloud giants and profitable chip manufacturers often possess the operational cash flow to fund their own growth. However, the risk is concentrated among companies that rely heavily on external financing to fuel their expansion.
Specialized cloud providers are currently facing significant capital expenditure gaps. For instance, CoreWeave reported nearly $2.1 billion in revenue against $6.8 billion in capital expenditures for the first quarter of 2026. Similarly, Nebius Group posted $399 million in revenue while spending nearly $2.5 billion on infrastructure.
The Funding Challenge
The issue is not a lack of demand. CoreWeave concluded the first quarter with a $99.4 billion revenue backlog, and Nebius held nearly $4.8 billion in deferred revenue. The core challenge lies in the “timing gap”: these companies must spend billions on GPUs and data centers to deliver on future revenue long before the actual cash hits their balance sheets.
Iren (NASDAQ: IREN), another key Nvidia partner, faces similar hurdles. While Nvidia has the option to purchase up to 30 million Iren shares at $70 each, the company is already dealing with heavy spending, having laid out $1.36 billion for hardware and property in the third quarter of fiscal 2026, despite generating only $144.8 million in revenue.
What Lies Ahead for AI Investors
Nvidia’s stake in these entities is only one piece of a much larger puzzle. The long-term viability of the AI sector depends on whether these specialized companies can eventually transition to self-funding their operations through business revenue. If the current flow of external financing dries up, companies heavily dependent on outside capital may be forced to significantly decelerate their expansion plans.