The Rundown: Deep Dive – Is Nvidia Funding a Data Center Collapse?

Date: September 27, 2025
Host: Zaid Admani, for Public.com

Brief Overview

In this weekend's deep dive, host Zaid Admani unpacks Nvidia’s rapidly expanding influence over the AI economy—not only as the leading supplier of AI chips but now as a central investor in its own customer base. The episode digs into Nvidia’s recent high-profile investments in AI infrastructure companies like CoreWeave and OpenAI, explores accusations of circular financing, and raises questions about sustainability and systemic risk as AI capex skyrockets. This nuanced analysis helps investors understand the difference between genius ecosystem building and risky, bubble-prone feedback loops.

Key Discussion Points and Insights

1. Nvidia’s Evolution: From Supplier to Investor

[00:00]

• Nvidia has been the "biggest beneficiary of the AI boom," becoming the "puppet master of the AI economy."
• After reaping massive profits from AI chip sales, Nvidia moves from "selling shovels in a gold rush" to "investing in the gold mine themselves."
• Raises concerns about possible "circular accounting."

2. The CoreWeave Partnership

[01:03]

• CoreWeave acts as a "landlord for AI chips," buying Nvidia hardware in bulk to outfit data centers, then renting computing power to customers like Microsoft and OpenAI.
• Nvidia holds about a 7% stake in CoreWeave and recently struck a $6.3 billion deal:
  • Nvidia agrees to buy unsold CoreWeave cloud capacity through 2032—a "massive safety net."
• Notable quote (Zaid Admani):

  "That investment that Nvidia made in CoreWeave will likely just come back to Nvidia because CoreWeave will use that money to buy Nvidia's chips." [02:36]

3. $100B Commitment to OpenAI

[03:09]

• Nvidia announced intent to invest up to $100 billion into OpenAI for new data centers powered by Nvidia chips.
• The investment will be delivered in $10B tranches as OpenAI builds out each gigawatt of capacity.
• This structure secures OpenAI as a long-term customer and further entangles Nvidia in the AI value chain.
• Notable quote (Zaid Admani):

  "It's another case of Nvidia giving a company money just to see that money come back to them." [04:17]

4. Surprising Intel Investment

[04:41]

• Nvidia’s $5 billion investment in Intel, a longtime competitor, seems politically motivated:
  • Supports the Trump administration’s push for Intel’s comeback.
  • Could lay groundwork for Intel eventually manufacturing Nvidia chips domestically, reducing reliance on Taiwan's TSMC.

5. Nvidia’s Cash Mountain and Shareholder Strategy

[05:44]

• Nvidia’s annual free cash flow soared from $6 billion to an estimated $97 billion in just three years—second only to Apple among tech giants.
• Major buybacks: $50 billion repurchased in the past year, another $60B authorized.
• Still, Nvidia is forced to seek alternative outlets for its cash, leading to significant equity investments in AI startups globally (over 50 deals in 2025 so far).
• Notable quote (Zaid Admani):

  "They're making $100 billion a year in cash. So they're using that cash pile now to invest in other AI companies." [06:12]

6. The Circular Financing Controversy

[07:07]

• Definition: Circular financing occurs when companies invest in customers, who in turn use that money to buy the investor’s own products.
• Analogy:

  "If Ford invested $10 million into a taxi company. But as part of that investment, the taxi company had to buy $10 million worth of Ford cars."

• Applies directly to Nvidia’s deals—e.g., OpenAI expected to use Nvidia funding to purchase Nvidia chips.
• Nvidia claims OpenAI will not use investment for “direct purchases,” but skepticism abounds.
• Some Bulls call this “genius financial engineering,” but critics see dangerous bubble risk.
• Notable quote (Zaid Admani):

  "Is Nvidia's growth moving forward real—or is it just propped up with their own cash?" [08:54]

7. Comparisons to Past Bubbles

[09:15]

• Historic analogs:
  • 1800s railroad boom
  • Late-1990s telecom bubble
• Massive infrastructure overbuild led to eventual crashes.
• Stats highlight market-wide dependence on AI capex:
  • Since ChatGPT’s launch, AI stocks drove 75% of S&P 500 returns.
  • AI = 80% of earnings growth, 90% capex (J.P. Morgan).
  • AI-related capex expected to exceed $300B this year and still growing.
  • Remarkable stat: "Construction spending on data centers has now eclipsed construction spending for corporate offices." [10:47]
• Problem: Capex outpacing cash flows; some contracts signed with “future money” and growing reliance on debt.
  • Bain & Co projects by 2030, there will be an $800B annual revenue gap between what’s needed to fund AI compute/builds and what AI companies are likely to earn.

8. Risks and Host’s Closing Take

[11:08]

• Market is “actively looking for reasons to be bearish,” with Nvidia’s wheel of investments a fresh red flag.
• On one hand, Nvidia’s approach is “admirable”—using cash to grow the AI ecosystem. On the other, it adds to fears of a self-fueling bubble.
• Vivid analogy:

  "They’re not just selling shovels anymore… they’re now bankrolling the miners, building the gold mine towns and taking a cut of all the gold that comes out of the ground." [11:22]

• Ultimate warning:

  "If the demand for AI dries up, that could cause the entire AI house of cards to come tumbling down." [11:27]

• Host’s view: Crash may not match dot-com era, but warning signs deserve investor attention.

Notable Quotes & Memorable Moments

• On circular financing:

  "That's kind of what's going on with these Nvidia deals… money just made a round trip." [07:35]

• On data center buildout:

  "OpenAI plans to use that money from Nvidia to build out their data centers and power them with Nvidia's chips." [03:26]

• On bubble risk:

  "The spending on AI that we’re seeing today is starting to feel like a lot of other historic bubbles." [09:15]

• On AI capex and the economy:

  "AI capex has actually added more to US GDP growth than consumer spending." [10:17]

Timestamps for Important Segments

| Timestamp | Segment | |-----------|---------| | 00:00 | Introduction to Nvidia as “puppet master” of AI economy | | 01:03 | CoreWeave partnership details | | 03:09 | Nvidia’s $100B investment in OpenAI | | 04:41 | $5B investment in Intel, political motives | | 05:44 | Nvidia’s skyrocketing free cash flow and buybacks | | 07:07 | Circular financing explained & controversy | | 09:15 | Historic bubble comparisons | | 10:17 | Data center and AI capex stats | | 11:08 | Host’s take: risks and analogies | | 11:27 | Ultimate warning on systemic AI risk |

Conclusion

This episode offers a concise yet comprehensive look at Nvidia’s pivotal role in both fueling and owning the AI infrastructure boom. The host lays out the mechanics of circular financing, analyzes the scale of Nvidia’s cash hoard and strategic risks, and draws stark parallels to previous tech infrastructure bubbles. Investors are cautioned to weigh whether Nvidia’s aggressive ecosystem investments represent “genius financial engineering” or troubling signs of a self-reinforcing financial bubble—one that could have far-reaching implications if growth ever falters.