NVIDIA doesn't want to sell chips for a lifetime.

Author | Ding Mao

Editor | Zhang Fan

On September 22nd, another significant piece of news emerged from the AI market.

Nvidia and OpenAI announced the signing of a letter of intent to establish a milestone strategic partnership. According to reports, OpenAI will deploy at least a 10GW data center based on Nvidia's computing power system for the training and operation of its next - generation models. To support this plan, Nvidia promised to invest up to $100 billion in OpenAI in stages.

The two parties expect that the first - stage capital injection will take place after the completion of the first GW data center in the second half of 2026, mainly based on Nvidia's next - generation Vera Rubin platform.

After the news was announced, the market responded positively. At the close of the U.S. stock market on Monday, Nvidia's stock price rose nearly 4%, and its total market value approached $4.5 trillion.

So, as ordinary investors, how should we view this investment? What far - reaching impacts will it have on Nvidia and OpenAI?

A Capital Game of "Chips for Equity"

In essence, this huge investment worth tens of billions of dollars is not simply a capital injection, but more like a deep strategic cooperation and resource exchange of "chips for equity" led by Nvidia.

Let's briefly explain the logic. From the news disclosed by the media, in this investment, Nvidia will invest a total of about $100 billion in OpenAI in stages according to the construction progress of the data center. The funds will mainly be used by OpenAI to build an AI data center with a total capacity of 10GW. A very important point is that each capital injection from Nvidia is deeply related to the progress of OpenAI's data center construction, ensuring the dedicated use of the investment funds.

After receiving Nvidia's investment funds in batches, OpenAI needs to purchase about 4 - 5 million GPUs and systems based on the new - generation Vera Rubin platform from Nvidia for the construction of the 10GW data center.

This back - and - forth forms a delicate capital closed - loop: Nvidia injects funds into OpenAI, and OpenAI purchases Nvidia's chips and systems through itself or its computing power partners. Eventually, the funds flow back to Nvidia. This is a capital game of moving money from one hand to the other, with the core being the transfer of assets rather than the consumption of cash.

So, in this capital game, what resources have the two parties exchanged?

For OpenAI, with the financial support of Nvidia, it directly obtains Nvidia's most advanced AI acceleration chips and builds an autonomous high - capacity data center, alleviating the huge future computing power demand gap. At the same time, based on the strategic partnership, OpenAI can also get in - depth software and ecological empowerment from Nvidia, enabling it to maintain its leading position in computing power support.

In contrast, through this investment, on the one hand, Nvidia locks in huge incremental chip sales revenue in the next few years. At the earnings conference call in August this year, Jensen Huang introduced that the cost of building a 1GW data center capacity is between $50 - 60 billion, of which about $35 billion is directly used to purchase Nvidia's AI chips and systems. Since the data center capacity that OpenAI plans to build this time is about 10GW, by analogy, the total construction cost is about $500 - 600 billion, of which about $350 billion is directly used to purchase Nvidia's AI chips and systems.

On the other hand, Nvidia obtained about 25% of OpenAI's equity through a $10 billion investment (calculated based on OpenAI's valuation of about $500 billion at the time of the news disclosure), and is expected to continuously share the stable incremental software revenue and potential AGI value brought by OpenAI in the future.

In addition, the more far - reaching significance of this investment is that Nvidia has deeply locked in OpenAI, a global top - tier AI player, sending a strong signal to the industry: at least in the foreseeable future, Nvidia's dominant position in AI infrastructure construction remains unshakable, greatly weakening the recent market concerns about "de - Nvidification".

For the industry, this also indicates that the competition in AI infrastructure has entered a new stage. It is no longer a simple hardware purchase and sales behavior between upstream and downstream as in the past, but has turned into a comprehensive competition around capital, technology, talent, and strategic alliances.

The "Exclusive" Alliance with Microsoft is on the Verge of Collapse

As OpenAI's most important partner and strategic ally in the past, Microsoft has invested more than $13 billion in OpenAI since 2019.

In return, on the one hand, OpenAI agreed to use Microsoft Azure as its sole cloud service provider. This enables Microsoft to provide the huge computing power required by OpenAI for training and running models through the Azure platform. As the top - tier model provider in the AI market, this investment is like betting on the most stable trump card in Microsoft's Azure growth blueprint.

On the other hand, through the investment, Microsoft obtained the right to use OpenAI's technology and can deeply integrate OpenAI's most advanced model technology into almost all its products, including Microsoft Office, Bing, and GitHub Copilot. The support of OpenAI's technology not only enables Microsoft's applications to make a leap from "tools" to "intelligent partners", upgrading the software business to a sustainable subscription - based revenue model, but also ensures Microsoft's leading position in the second half of the AI application competition.

From this perspective, it is precisely based on the in - depth alliance with OpenAI that ensures the explosive growth of Microsoft Azure and the reshaping of product strength brought by AI's in - depth empowerment of software, which is also the fundamental reason for Microsoft's continuous favor in the capital market.

Although this win - win alliance seems unbreakable, OpenAI is not without other thoughts. The previous cooperation with Oracle was an important challenge to the stability of the alliance, and with Nvidia's entry, the "exclusive" alliance between Microsoft and OpenAI may be facing an accelerating risk of collapse.

The underlying logic is that Nvidia directly built a "computing power fortress" independent of Azure for OpenAI through the way of "chips for equity", helping it get rid of future computing power constraints, reducing its dependence on Microsoft Azure, and enhancing its strategic independence and bargaining power.

More importantly, if OpenAI is no longer restricted by Microsoft in terms of computing power, it will directly weaken Microsoft's ability to control and influence OpenAI through cloud services. This means that in the future, OpenAI may no longer serve as the exclusive "brain" of Microsoft's software products, but will transform into an important ally relationship. As an independent third - party, its model technology can empower not only Microsoft but also other giants such as Google and Amazon, further expanding its revenue sources.

Once this expectation comes true, the market pattern of future AI application competition may change fundamentally. Microsoft's differential advantages in the AI application field will face challenges, and its valuation logic may also be re - examined by the capital market.

Re - evaluating Nvidia

For Nvidia, in addition to the direct performance benefits and the consolidation of its leading market position, the more important thing about this investment is that it may trigger Wall Street to re - evaluate its long - term value.

In the upward cycle of artificial intelligence, Nvidia has been extremely prominent as an absolute overlord. However, in fact, its valuation is still essentially limited by its identity as a "hardware company": its revenue growth is extremely dependent on a large amount of capital investment, with strong periodicity and a heavy - asset model.

In contrast, the valuation logic of software companies represented by OpenAI is completely different. Although the huge early R & D and computing power investment squeezes profits to a certain extent, once the scale effect is formed, its revenue is highly scalable. Subscription revenue will show exponential growth, and cash flow is also more stable. At the same time, the extremely low marginal cost also means more considerable gross profit margins, making the capital market willing to give it a higher valuation premium.

Through this strategic investment, Nvidia has successfully deeply bound itself to the future of general artificial intelligence (AGI). In addition to the huge revenue provided by its core hardware business, Nvidia also continuously shares the potential incremental revenue with higher growth, better profitability, and more stable cash flow brought by future software business by holding OpenAI's equity, ensuring its possibility of continuous profit - making at all stages of AI development.

This means that the valuation logic of Nvidia in the capital market may have undergone a qualitative change. That is, it has transformed from a traditional upstream core AI chip supplier to a super - dominant player of "AI hardware + ecosystem + software services" in the future. Through diversified layout, it can effectively hedge the risk of growth stall in different stages of AI.

There is no lack of precedents for this kind of valuation logic switch on Wall Street.

In the past, as a hardware - dominated company, Apple's valuation fluctuated with the iPhone sales cycle. In recent years, with the increase in the proportion of software services, Apple's valuation logic has changed from a simple hardware device brand to an "integrated hardware and software" ecosystem service provider. It is precisely based on this switch in valuation logic that has greatly enhanced the stability and sustainability of Apple's market value, enabling it to get rid of the cycle constraints of smartphone hardware to a certain extent and have a more stable independent market trend.

In the future, with the implementation of this strategic investment, a valuation reshaping journey for Nvidia may be about to begin.

Chart: Business transformation promotes the reshaping of Apple's valuation logic. Data source: wind, compiled by 36Kr

*Disclaimer:

The content of this article only represents the author's views.

The market is risky, and investment should be made with caution. Under no circumstances do the information in this article or the opinions expressed constitute investment advice to anyone. Before making an investment decision, if necessary, investors must consult professionals and make careful decisions. We have no intention to provide underwriting services or any services that require specific qualifications or licenses for the parties involved in the transaction.