OpenAI has triggered another wave of significant stock surges. This marks the fourth occurrence in just over a month. Companies that align themselves with OpenAI's momentum see their stock prices soar. Within a single month, OpenAI has secured 26 gigawatts of computing power from NVIDIA, Advanced Micro Devices, and Broadcom.

**Broadcom's Double Stock Surge**

Semiconductor giant Broadcom closed Monday with a remarkable 10% gain, pushing its market capitalization close to $1.7 trillion and surpassing Taiwan Semiconductor Manufacturing Company to become the second-largest semiconductor company by market value, trailing only NVIDIA. Under the leadership of CEO Hock Tan, Broadcom has benefited substantially from the AI boom, with shares doubling this year.

The latest catalyst for Broadcom's stock surge is once again the wealth-generating phenomenon that is OpenAI. This generative AI giant has not only achieved a valuation exceeding $500 billion, making it the world's highest-valued startup, but has also earned a reputation as having the "golden touch." Any company that strikes a deal with OpenAI subsequently experiences stock price spikes.

Broadcom's surge stems from announcing a strategic partnership with OpenAI, where both companies will jointly deploy custom 10-gigawatt (GW) class AI chips. OpenAI will handle AI chip design and collaborate with Broadcom on development and deployment. Unlike previous deals with NVIDIA and Advanced Micro Devices, the OpenAI-Broadcom partnership doesn't involve investment or subsidies.

According to joint press releases, OpenAI will integrate its cutting-edge model and product development expertise directly into hardware products through these self-developed chips and systems, unlocking new computational capabilities and intelligence levels. The project is expected to begin deployment and network system refinement in the second half of 2026, with completion targeted by the end of 2029.

This actually marks Broadcom's second stock surge attributed to OpenAI. Last month, after Broadcom's earnings report, CEO Hock Tan announced a mysterious new customer had committed to $10 billion in new orders. Markets speculated this customer was OpenAI, driving Broadcom shares up 16%.

Now that both companies have formally announced their collaboration, despite this being anticipated news, investors remain excited, propelling Broadcom's stock to a second wave of gains. Tan has maximized OpenAI's stock-boosting effect, truly achieving a "double benefit from one fish."

**Following OpenAI Guarantees Gains**

Over the past month, OpenAI has continuously announced several major deals worth hundreds of billions of dollars, securing 26 gigawatts of computing power from NVIDIA, Advanced Micro Devices, and Broadcom. CEO Sam Altman has consistently driven up cloud service providers' and chip giants' stock prices while securing optimal partnership terms for his company.

Last month, OpenAI announced a $300 billion cloud services contract with Oracle spanning five years. Oracle's stock skyrocketed 36% in a single day, marking its largest gain in over three decades, with market capitalization surging by $250 billion and briefly exceeding $900 billion.

Several weeks ago, OpenAI and NVIDIA announced a strategic investment agreement. NVIDIA plans to invest up to $100 billion in OpenAI, supporting the company's purchase of NVIDIA AI processors to build data center clusters costing hundreds of billions of dollars. On the announcement day, NVIDIA's market value jumped by over $200 billion, pushing total market capitalization beyond $4.5 trillion.

A week ago, OpenAI and Advanced Micro Devices announced strategic cooperation: OpenAI will deploy a total of 6 gigawatts (GW) of Advanced Micro Devices processors over the coming years, with the first 1-gigawatt deployment expected to begin in the second half of 2026, utilizing Advanced Micro Devices' upcoming MI450 series. This news drove Advanced Micro Devices shares up 35% last week, with market capitalization soaring past $350 billion.

In just over a month, OpenAI has consecutively concluded four major partnership deals covering cloud infrastructure, AI processor procurement, and collaborative development of proprietary chips. This not only secures future computing power but also achieves processor supplier diversification while planning to build proprietary AI infrastructure.

Altman's ambitions extend beyond this. In an August media interview, he envisioned OpenAI's expansion into consumer hardware, brain-computer interfaces, and social media. He stated that OpenAI would spend trillions of dollars on data center construction in the "near future" and hinted at interest in acquiring Chrome browser if the U.S. government forces Alphabet to sell it.

Altman aims to transform OpenAI into a full-stack AI giant similar to Alphabet, becoming a conglomerate spanning large language models, consumer hardware, processors, social platforms, and data centers. If these plans materialize, OpenAI's valuation could multiply several times, reaching trillions of dollars and rivaling Alphabet, Microsoft, and Apple.

**Even Musk Shows Skepticism**

However, behind this AI great leap forward lurks another risk: OpenAI has signed deals worth hundreds of billions of dollars, but as a startup, do they actually have the money to pay and implement these agreements? If lacking sufficient funds while continuously signing undeliverable mega-deals to boost valuations in capital markets and AI company stock prices, does this AI boom conceal a massive bubble?

Citigroup analysts estimate that each additional gigawatt of computing capacity requires approximately $50 billion in infrastructure investment. Therefore, OpenAI's latest partnerships with chip manufacturers are expected to generate cumulative capital expenditures of approximately $1.3 trillion by 2030.

OpenAI is undoubtedly the strongest fundraising startup in history, with their funding rounds and valuations continuously breaking venture capital records. Just this April, they completed a record-breaking $40 billion funding round, reaching a $300 billion valuation, with SoftBank's Masayoshi Son leading $30 billion and Microsoft continuing to participate. Of this, $10 billion was secured by mid-April, with the remaining $30 billion expected by year-end.

However, SoftBank's total investment will be tied to OpenAI's for-profit transformation. If OpenAI's parent company cannot successfully transition to for-profit status, SoftBank's investment will decrease to $20 billion. This explains why Musk continues using litigation and acquisition attempts to prevent OpenAI's parent company transformation, seeking to create catching-up time for his xAI.

In less than six months, OpenAI's latest valuation has grown to an astounding $500 billion, surpassing Musk's SpaceX to become the highest-valued startup in history. Musk can only somewhat sourly comment on X that OpenAI's valuation is excessive.

While Musk may harbor animosity, OpenAI's frenzy faces numerous questions. In just over a month, the massive investments and partnerships between OpenAI, NVIDIA, Advanced Micro Devices, Oracle, and Broadcom have triggered the largest capital storm of the AI era.

With such large-scale computing power investments, whether OpenAI can generate corresponding commercial returns remains questionable. When an unprofitable startup leverages hundreds of billions in valuation to move global capital markets and weaves a network of capital and technology-bound transactions with chip giants and cloud service providers, the industry naturally debates whether we're witnessing necessary infrastructure for a technological revolution or observing a dangerous capital circulation game.

**Revenue and Funding Cannot Support Burn Rate**

Indeed, OpenAI's revenue growth is quite remarkable. Their revenue primarily comes from ChatGPT subscription revenue, enterprise API payments, and selling customized AI solutions to companies like Microsoft. While current revenue mainly derives from user subscriptions with weekly active users reaching 800 million, they're actively transitioning toward an enterprise AI infrastructure-focused revenue model.

Following last year's 150% revenue growth, OpenAI's annualized revenue in the first half of this year exceeded $10 billion, with full-year revenue expected to grow over twofold to $12.7 billion. According to current plans, OpenAI projects 2030 revenue of $200 billion.

However, such dramatic funding amounts and rapid revenue growth cannot satisfy OpenAI's cash-burning appetite. Last year, OpenAI posted net losses of $4 billion, expecting this to double to $8 billion this year. To maintain advantage in the AI arms race, OpenAI pledges to invest over $115 billion over five years, primarily for self-built data centers and computing power investments.

This burn rate far exceeds OpenAI's revenue and fundraising pace. Altman must simultaneously negotiate partnerships and raise funds. To minimize capital requirements, Altman adopted innovative "investment-procurement" models for NVIDIA and Advanced Micro Devices processor purchase agreements.

NVIDIA invests up to $100 billion in OpenAI equity, while OpenAI uses these funds to purchase NVIDIA processors. For Advanced Micro Devices purchases, after OpenAI procures sufficient processors and drives Advanced Micro Devices sales and stock price growth to targets, Advanced Micro Devices will provide OpenAI with up to 10% warrant coverage, essentially offering performance rewards. Media reports indicate Altman also sought investment from Taiwan Semiconductor Manufacturing Company, but they declined.

While NVIDIA GPUs can be offset through equity investment, Advanced Micro Devices GPU purchases and Broadcom's self-developed chips require substantial upfront capital investment. Moreover, OpenAI's cloud infrastructure services contract with Oracle carries an annual bill of $60 billion. Where does OpenAI get the money to pay?

Oracle appears unconcerned. Oracle Co-CEO Clay Magouyrk expressed confidence this week that OpenAI "certainly" can pay the annual $60 billion cloud resource fees. He mentioned that OpenAI's user growth rate is unprecedented, and AI technology will have tremendous impact across all industries and enterprise types.

NVIDIA CEO Jensen Huang also clarified external questions about "supplier financing." He stated that OpenAI temporarily lacks money and needs to raise funds based on future exponential growth prospects. So they gave us the opportunity to invest alongside other investors. Huang even expressed regret at not having more money to invest in OpenAI. He believes OpenAI will become a trillion-dollar company, while NVIDIA will become the first company with over $10 trillion market capitalization.

Interestingly, while external observers worry about capital bubbles in OpenAI's aggressive expansion, Altman himself warns of AI industry bubble risks. He publicly stated over a month ago that he also believes the AI market is in a bubble period, but AI remains the most important innovation in a long time.

**Reminiscent of the Dot-Com Bubble**

The current AI investment frenzy reminds many of the internet stock bubble over twenty years ago. At the turn of the century, investors frantically poured into emerging internet companies, believing the internet would change all business models, blindly driving up market values and valuations of internet companies lacking profitability and business models.

Over subsequent decades, the internet indeed transformed nearly all previous industries, but this didn't prevent bubble burst and market collapse. From March 2000 to October 2002, the Nasdaq index lost nearly 80% of its value, with many internet companies failing to survive.

Many industry insiders have mentioned bubble risks in current AI field investments. Alibaba co-founder Joe Tsai stated in March that many large tech companies, investment funds, and other institutions are competing to build AI training server bases, with this trend appearing somewhat blind. Many projects lack clear customers during construction.

While Tsai believes the AI market could reach at least $10 trillion, he emphasizes AI isn't a winner-take-all field, with true victory lying not in developing the most powerful AI models but in faster technology implementation and application. Tsai and Jack Ma personally experienced the dot-com bubble burst.

Apollo Global Management Chief Economist Torsten Slok believes today's AI bubble is actually larger than the internet bubble, with the S&P 500's top ten companies having higher forward price-to-earnings (P/E) ratios today than the top ten companies during the mid-1990s tech bubble.

In his view, the current AI bubble differs importantly from the internet bubble: the numerous internet startups of that era dispersed bubble risks, while current market risks are highly concentrated in very few tech giants, with valuations even more exaggerated than before.

But optimists always exist. Constellation Research Chief Analyst Ray Wang believes, "From a broader AI and semiconductor investment perspective, I don't think this is a bubble. Supply chain fundamentals remain strong, and AI trends' long-term trajectory supports continued investment."

Bank of America Securities Managing Director Vivek Arya notes that while AI companies are locked in a race to protect dominance or find new revenue streams, today's tech giants have much stronger financial strength than internet companies of the past.

Top cloud service providers now have sufficient operating cash flows to support their capital expenditures. Compared to internet companies of the dot-com era that always relied on debt financing, current top cloud service providers have capital expenditure intensity of about 25%, while expected operating cash flows average over 30%.

NVIDIA currently trades at 51 times earnings, while 2026 expected P/E is only 29 times, "far below" its earnings growth rate. Analysts state this is "completely different" from Cisco, Nortel Networks, and Yahoo stocks trading at 100 times valuations twenty years ago.

More importantly, the Federal Reserve is currently in a rate-cutting cycle, starkly different from the Fed's continuous rate hikes during the 2000 and 2008 market crashes.