A new AI supercycle is beginning HSBC expects OpenAI to require up to US$207 billion in new financing by 2030

The Zhitong Finance App learned that HSBC Global Investment Research recently released a report saying that after updating OpenAI's computing power requirements and cloud rental cost forecasts, it is estimated that this artificial intelligence unicorn may require up to 207 billion US dollars in new financing by 2030. The research team, led by Nicolas Cote-Colisson, pointed out that OpenAI recently significantly added long-term procurement agreements with cloud vendors, driving a surge in funding requirements.

According to HSBC analysis, OpenAI has announced that it has signed a cloud computing purchase agreement totaling 250 billion US dollars with Microsoft (MSFT.US) and another 38 billion US dollar cloud service procurement contract with Amazon (AMZN.US) for a period of seven years. At the same time, competitor Anthropic (supported by Amazon and Google) is also speeding up expansion, announcing that it will obtain the supply of 1 million TPU AI chips from Google (GOOG.US, GOOGL.US), worth “tens of billions of dollars,” and plans to invest 50 billion US dollars in infrastructure construction. At the same time, it has reached a 30 billion US dollar computing power commitment agreement with Microsoft and Nvidia (NVDA.US), which is expected to receive up to 15 billion US dollars in capital injection.

The Cote-Colisson team said that after re-evaluating OpenAI's computing power expansion path and leasing costs, they came to the conclusion that it would still require an additional 207 billion US dollars by 2030. Whether the financing gap can be reduced depends on a number of unknown variables, including whether OpenAI can adjust procurement commitments according to actual needs, and whether future pressure can be relieved through equity injections, debt financing, or significantly higher than expected revenue.

The report points out that a new AI supercycle is beginning, and OpenAI is expected to bear 1.4 trillion US dollars in computing power costs over the next eight years, which has undoubtedly heightened investors' concerns about its return on investment. Compared to the 2025 revenue estimate of about US$12.5 billion, the huge gap between cost and revenue also affects the risk assessment of the entire AI industry chain.

However, analysts pointed out that the CEO of Microsoft has stated that AI model companies, infrastructure providers, and chip makers simultaneously promote commercial implementation, which will help enterprise customers achieve AI value faster. Therefore, even without considering AGI (General Artificial Intelligence), there is still huge room for productivity improvement brought about by AI.

Among the companies covered by HSBC, the most vulnerable to the success or failure of OpenAI include Oracle (ORCL.US), Microsoft, Amazon, Nvidia, AMD (AMD.US), and SoftBank, which is deeply committed to AI.