--- title: "Fu Peng's latest thoughts: OpenAI - Oracle - NVIDIA's trillion-dollar investment, under the AI capital expenditure frenzy, the pricing logic and risk concerns of BTC's computing power【Fu Peng's article】" description: "Fu Peng explored the pricing logic of BTC's computing power in his latest article, believing that BTC is gradually becoming a financial derivative of computing power prices. The article analyzes the i" type: "news" locale: "en" url: "https://longbridge.com/en/news/259371276.md" published_at: "2025-09-29T14:35:46.000Z" --- # Fu Peng's latest thoughts: OpenAI - Oracle - NVIDIA's trillion-dollar investment, under the AI capital expenditure frenzy, the pricing logic and risk concerns of BTC's computing power【Fu Peng's article】 > Fu Peng explored the pricing logic of BTC's computing power in his latest article, believing that BTC is gradually becoming a financial derivative of computing power prices. The article analyzes the impact of computing power supply and demand and costs on BTC prices, pointing out that when there is a shortage of computing power, BTC prices and mining company stock prices will rise rapidly, while an oversupply of computing power may lead to a wave of mining company bankruptcies. Fu Peng emphasized that the high cost of computing power has stimulated the expansion of hardware investment, and the expected supply of computing power will continue to increase in the future "Looking at the world from the trading desk, Eric Wu discusses finance" **"Eric Wu Speaks · Season 6" is fully upgraded! Early bird countdown: 2 days! Subscribe now** ## BTC: The Price of Computing Power During this time, after communicating with many people in the artificial intelligence industry chain, I increasingly feel that BTC is just a price of computing power. Its production method is entirely a process of computing power, and unlike the industry chain, it does not have a final application. Therefore, it gradually becomes a symbol of value, what value? Computing power. However, it is not as purely a credit hedge and faith value as gold; it has a production mechanism closely related to computing power itself. So, does it increasingly represent a financial derivative of computing power prices? If this thinking is correct, then we need to consider what factors in the computing power segment will trigger a decline in computing power prices and an increase in computing power prices, with shortages leading to increases, excess leading to decreases, high costs leading to increases, and low costs leading to decreases. For example, the upper part of the image shows the price trend of BTC since its inception (logarithmic scale), and the lower part shows the volatility of BTC (blue line) along with an example of a mining company (Riot Platforms (RIOT.US)) stock price. Mining companies are actually very similar to BTC's volatility. Of course, since mining itself is upstream, **when computing power is scarce and costs are high, once BTC's volatility amplifies, it will rapidly rise, and the upstream mining companies will also rise rapidly.** However, in the past two years, it feels like the high computing power prices have low volatility at high prices, while mining companies have remained stable. Lower volatility means price maintenance, and stock prices can only maintain and rise. This raises the question: if volatility amplifies and computing power prices begin to decline with increased capital expenditure and investment, the amplification of volatility may lead to a wave of mining company bankruptcies. If the attributes of BTC we discussed earlier are correct (that is, it is a price of computing power determined by the supply and demand of computing power and costs), then according to the general rules of industrial development, the first stage is the high cost of computing power, stimulating an expansion of upstream hardware investment and capital expenditure. The price of computing power and investment in computing hardware create positive feedback; the more expensive it is, the more valuable hardware investment becomes. As capital expenditure continues to expand, future expectations for computing power supply will increase (this stage is a good thing for NVDA, the shovel seller). At this point, if the demand for computing power (application side) has not fully kept up (or if high computing power costs prevent applications from quickly catching up), changes in the supply and demand curve will impact computing power prices. Perhaps the bifurcation of BTC and NVDA is reflecting this stageConsidering the recent news about the concerns of many investment giants on Wall Street regarding the enormous capital expenditures in artificial intelligence, it is indeed necessary to contemplate the potential stages that this industry development will inevitably go through. ## How to Understand the Mutual Investments Among AI Giants? The key moment for OpenAI in 2023 can be said to have ignited the development of artificial intelligence, especially as people began to envision how general large models and data could be transformed into practical applications. Therefore, starting from 2023, this trend towards application has driven upstream capital expenditures and infrastructure construction, such as data centers and upstream hardware. The rapid rise of OpenAI is also reflected in the significant growth of upstream NVIDIA's stock price and revenue; I previously wrote a diary describing NVIDIA as the one selling shovels during the gold rush. By now, there is no doubt that upstream data centers have become the core of a new round of large-scale infrastructure. The world is accelerating the promotion of this early infrastructure project that concerns future productivity. However, at present, after investors have tasted the sweetness of upstream infrastructure, they are facing a critical question: whether the current enormous capital expenditures could bring significant risks. Especially recently, NVIDIA signed a cooperation agreement with OpenAI to jointly build data centers. According to the project's implementation progress, NVIDIA will gradually invest up to $100 billion in OpenAI. Meanwhile, OpenAI is spending $300 billion (over five years) to procure computing power resources from Oracle, which in turn is spending $40 billion to purchase graphics cards from NVIDIA. As this cash flow cycle of "you invest in me, I invest in you, you buy from me, I buy from it" begins to emerge, similar concerns have also started to appear on Wall Street. It's akin to the sellers of shovels during the gold rush making substantial profits, but with too much cash that cannot be further utilized, leading to an excess of cash flow on the books. If they invest by expanding production capacity, it may trigger the risk of supply exceeding demand, resulting in a decline in gross margins. For companies that hold absolute monopolies in gold rush towns, the urgent task is not to increase production capacity but to expand demand. We see Jensen Huang stating that they primarily arrange orders based on customer demand and will not blindly expand production capacity beyond demand. In fact, this statement also indicates that in a monopolistic environment, he fully understands the core value and maintains the current gross margin, allowing demand to grow rapidly, which is the most beneficial interest for NVIDIA. Therefore, the sellers of shovels can allocate substantial funds to directly support customers in purchasing their own equipment. If customers have strong profitability in the future, it can not only bring high returns but also establish an absolute monopoly position. For instance, OpenAI can be viewed as a gold mine, Oracle as the mining team, while NVIDIA provides the core "gold shovel." This kind of business behavior is actually quite common and universal in many industries. In many construction companies, they often subsidize their cash flow to end customers. For example, in the real estate industry, we can compare real estate buyers to OpenAI, while Oracle is similar to real estate developers or construction companies, and steel mills or building material suppliers (like glass manufacturers) play a role similar to NVIDIA. Despite the different industrial characteristics, in the business model, the mutual investments among giants lock in monopolies, allowing for rapid expansion and fragmentation through this chainAll cash flows are concentrated on the demand side, further strengthening the oligopoly, possessing advantages that cannot be matched by separate decentralized calibration. OpenAI, Oracle, and NVIDIA have formed the first powerful alliance of complementary capital and technology through mutual order locking and investment, but all of this relies on the continuous expansion of demand; otherwise, the bundling relationship will amplify risks. Once the downstream application demand represented by OpenAI is insufficient, Oracle and NVIDIA will face order reductions and investment losses. **Similar situations have just been experienced in the Chinese real estate market:** Once cash flow turnover slows and debt issues arise, upstream construction companies (such as Nantong enterprises bound to Evergrande) or building material suppliers (such as Guangzhou-Foshan decoration companies) will be firmly trapped. ***“From a business perspective, it is a strategic alliance aimed at building a technological competitive advantage, but from a financial perspective, it is a leveraged behavior, just as Cao Cao enhanced stability with iron chains connecting boats, yet faced the hidden danger of 'burning the connected boats'”*** I believe there is no doubt about the great opportunity of artificial intelligence itself, and all parties will naturally voice their positions. For example, practitioners in the application sector will emphasize the enormous potential in this field, while upstream data centers and equipment manufacturing companies will certainly stress their significant demand. The current global industrial chain is basically clear, and the hot development of artificial intelligence infrastructure investment will inevitably continue, with broad prospects for the future. However, for investors in the secondary stock market, **the key to grasping the investment rhythm lies in the "capital market's" perception of events. This perception will affect the market, and then through the process of proving or disproving, a feedback mechanism will form. Therefore, the core at this stage may not be the already established argument that the artificial intelligence industrial chain will have great achievements in the future, but rather to pay more attention to the mainstream attitudes and arguments on Wall Street, whether to continue investing vigorously or to have concerns about it;** If we continue to focus on capital expenditure, then AI infrastructure will advance vigorously; if we focus on AI cost reduction and efficiency improvement, that is also positive and proactive; but if we focus on whether the demand application side can keep up with and meet the upstream's rapid pace that has already accounted for overly high expectations, then this may generate Wall Street's concerns this week about "huge capital expenditure investments." Therefore, we must focus on the demand side of the application field: if the demand is real and exceeds supply (continuously proven), this iron chain of boats will form a huge demand cycle; if the demand is disproven, it will become a fire that amplifies the overall volatility of the industrial chain; the only thing we need to closely monitor is the growth rate of the application side represented by OpenAI. Back to the old topic, no matter how everyone sells shovels, the prosperity of the gold town must ultimately come from the continuous ability to dig up gold. ## Fu Peng's Column Community This year marks the tenth anniversary of Wall Street's "Fu Peng Says" annual column! The early bird countdown for "Fu Peng Says Season 6" has only 2 days left! It ends at the end of September! Season 6 of Fu Peng Says has been fully upgraded, adding notes from Fu Peng's articles and chart interpretations based on daily videos. In addition, the class representative will also establish a **exclusive community** for purchasing users. Besides the daily content of Fu Peng Says, we will also send daily trading focus morning reports, noteworthy research reports, recommended electronic reading lists, extended reading materials, and everyone can ask questions and communicate more in the community to progress togetherUsers who join the column, please scan the QR code in the image below to add the assistant and enter the exclusive group chat for "Fu Peng Says." Risk Warning and Disclaimer The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investing based on this is at your own risk ### Related Stocks - [OpenAI.NA - OpenAI](https://longbridge.com/en/quote/OpenAI.NA.md) - [NVDA.US - NVIDIA](https://longbridge.com/en/quote/NVDA.US.md) - [ORCL.US - Oracle](https://longbridge.com/en/quote/ORCL.US.md) ## Related News & Research | Title | Description | URL | |-------|-------------|-----| | India’s top telco tackles AI with $110 billion build plan and proven fast market dominance playbook | India’s top telco, Reliance Jio, plans to invest $110 billion in AI infrastructure over seven years to enhance its servi | [Link](https://longbridge.com/en/news/276406785.md) | | 10:27 ETAmigos For Kids Selected as OpenAI Ready Award Recipient Through People-First AI Fund | Amigos For Kids has been selected as a recipient of the OpenAI Ready Award through the People-First AI Fund, which suppo | [Link](https://longbridge.com/en/news/276142228.md) | | Bartronics India Denies Collaboration With OpenAI | Bartronics India Ltd :COMPANY DENIES LAUNCHING 500 MW DATA CENTRECOMPANY DENIES COLLABORATION WITH OPENAI | [Link](https://longbridge.com/en/news/276327596.md) | | OpenAI finalizing funding round, 90% coming from strategic investors, CNBC says | OpenAI finalizing funding round, 90% coming from strategic investors, CNBC says | [Link](https://longbridge.com/en/news/276503935.md) | | OpenAI, Reliance partner to add AI search to JioHotstar | OpenAI has partnered with Reliance to integrate AI-powered conversational search into JioHotstar, allowing users to sear | [Link](https://longbridge.com/en/news/276347539.md) | --- > **Disclaimer**: This article is for reference only and does not constitute any investment advice.