Nvidia, Oracle, and AMD are just the beginning? Altman: OpenAI will have more big deals.

```

Is the "AI closed-loop economy" going to spiral even further?

Recently, a series of sensational agreements that OpenAI has struck with NVIDIA, Oracle, and AMD may be just the beginning. On October 8, according to TechCrunch, OpenAI CEO Sam Altman revealed on a blog show:

In the coming months, you can expect us to bring more (such deals).

He explained that, given the company’s strong confidence in the capabilities of its future models and the demand they will stimulate, “We have decided that it is time to make a very aggressive infrastructure bet.”

Almost simultaneously with Altman’s remarks, NVIDIA CEO Jensen Huang expressed surprise during a CNBC show at OpenAI’s partnership with rival AMD. When asked if he had prior knowledge of the deal, he replied he was “not really aware” of it.

Since the start of this year, OpenAI has finalized a series of agreements, including the $500 billion “Stargate” project with Oracle and SoftBank, as well as massive compute power collaborations with NVIDIA and AMD. The characteristics of these deals are innovative financing structures, described by outsiders as the prototype of an “AI closed-loop economy” or “AI chaebol.”

This strategy has a direct impact on the structure of the AI industry chain. By turning suppliers into stakeholders, OpenAI not only secures the massive computing power needed for future development but also tightly binds the fate of companies like NVIDIA and AMD to its own growth, increasing the systemic interconnectedness and risk of the entire ecosystem.

“Cross-shareholding”: Using suppliers’ money to buy the future

The cooperation model between OpenAI and the two major chip giants highlights its savvy transaction structure design. The core of these arrangements is leveraging OpenAI’s huge computational needs in exchange for deep binding and financial backing from chip suppliers.

In the partnership with NVIDIA, NVIDIA promised to invest up to $100 billion in OpenAI, becoming a shareholder in this AI company. In return, OpenAI will directly purchase AI equipment including GPUs from NVIDIA, preparing to become a “self-hosted hyperscale enterprise.” It is said that this model is referred to by some critics as AI “circular financing,” i.e., NVIDIA is essentially lending OpenAI the funds to buy its own products.

The deal structure between OpenAI and AMD is also quite special. Wallstreetcn mentioned AMD has agreed to provide OpenAI with a large number of stock warrants, totaling up to 10% of the company's shares, which will be gradually exercised over the next few years depending on the rise in stock price and other factors. In exchange, OpenAI will use and help develop AMD’s next-generation AI GPU chips.

This means OpenAI will become a shareholder of AMD, and its chip procurement costs will be partially offset by AMD’s own stock appreciation. UBS analyst Timothy Arcuri pointed out in a research report that if AMD’s stock price reaches certain milestones, the value of OpenAI’s holdings could be as high as $100 billion, enough to pay for most of its GPU procurement bills. Arcuri believes a more likely scenario is that OpenAI will sell AMD stock along the way to pay the bills; essentially, AMD is providing financing for its customer’s procurement.

Trillion-dollar Gamble: Altman's Aggressive Bet on AI's Future

OpenAI’s expansion is unprecedented in scale. Its deal list includes: the $500 billion “Stargate” project with Oracle and SoftBank aimed at building 10GW of US data center facilities; at least 10GW of AI data centers in collaboration with NVIDIA; and a 6GW computing power deployment with AMD.

What supports this string of bold deals is Altman's unwavering confidence in the future of AI. On the podcast he explained that, foreseeing future models’ capabilities and demand will far exceed today’s, the company “has decided it is time to make a very aggressive infrastructure bet.”

However, the reality is that OpenAI's financial situation is far from matching the scale of these deals. Reportedly, the company’s first-half 2025 revenue is $4.5 billion, yet the total value of contracts signed this year is estimated at as much as $1 trillion. NVIDIA CEO Jensen Huang also admitted that OpenAI “does not have the money for all this equipment yet,” estimating that each gigawatt of AI data center will cost $5 to $6 billion.

Despite a massive financial shortfall, Altman has shown enormous confidence in the future. He said, “I have never been so confident in the research roadmap ahead of us and the economic value that can be generated by using these (future) models.” At the same time, he admitted that achieving this goal would require support across the entire industry.

On the podcast, he said: “From the electronics level to model deployment and everything in between, we need the whole industry or a large part of it to support this. Therefore, we will work with a lot of people.”

These words explain why OpenAI is seeking such diverse and in-depth partnerships, and signal that future deals will touch ever-wider areas from energy to software distribution. For the tech industry, OpenAI’s next megadeal is something to watch with bated breath.

Risk warning and disclaimerThe market has risks, investments should be cautious. This article does not constitute individual investment advice, nor does it take into account the particular investment objectives, financial situation, or needs of any individual user. Users should consider whether any opinions, views, or conclusions in this article fit their specific circumstances. All investments made accordingly are at your own risk. ```