Openai and Oracle shocked the market this week $300 billion in five-year agreementwhich is part of a surge in new businesses that have caused cloud providers to soar. But maybe the market shouldn’t be surprised. This deal reminds you that despite Oracle’s legacy position, the company still plays a major role in the AI infrastructure.
On the Openai side, the protocol is more obvious than the lack of details suggests. First, startups are willing to pay so much for calculations to provide a measure of their appetite for startups—even if it is unclear where the calculations on electricity are said to come from or how it will be paid.
Chirag Dekate, vice president of research firm Gartner, told TechCrunch it was obvious why both parties were interested in the deal. He pointed out that it makes sense to work with Openai and work with multiple infrastructure providers. It also diversifies the company’s infrastructure (expanding risks among several cloud providers) and gives OpenAI an expanded advantage over its competitors.
“Openai seems to be putting together one of the most global AI supercomputing foundations to scale with proper inference,” Dekate said. “It’s very unique. It could be a model for the appearance of a model ecosystem.”
Compared to cloud competitors like Google, Microsoft Azure and AWS, some industry observers were surprised by the opinions of Oracle’s involvement. But DeCart believes observers should not be surprised: Oracle has previously worked with Hyperscalers and provides the infrastructure for Tiktok’s massive U.S. business.
“Over the past few decades, they have actually built core infrastructure capabilities that allow them to take extreme scale and performance as a core part of their cloud infrastructure,” Dekate said.
Payment and Power
But even as the stock market celebrates the deal, key details are missing and there are still issues with electricity and payments.
TechCrunch Events
San Francisco
|
October 27-29, 2025
Over the past year, Openai has issued a series of infrastructure investment announcements, and the prices of each infrastructure are eye-catching. Openai has committed to spending approximately $60 billion a year on Oracle’s computing and $10 billion Use Broadcom to develop custom AI chips.
Meanwhile, Openai said in June Annual recurring revenue of $10 billionFrom $5.5 billion last year, it reached about $5.5 billion. This figure includes revenue from the company’s consumer products, CHATGPT commercial products and its APIs. And its CEO, Sam Altman Rose colored pictures that paint their future prospects In terms of subscribers, products and revenue, the company burns through billions of dollars in cash every year.
Power is another issue, or more specifically, where the company plans to procure the energy required to run this level of calculation.
Industry observers have been predicting recent enhancement of natural gas Solar and batteries It can be said that in many markets, electricity can be released faster at a lower cost. So are technology companies Betting core.
Despite the headlines moving, the energy impact of Openai’s expected growth was not entirely unexpected. According to A Report Yangfang Group published yesterday.
Computing has always been a limit for AI companies, so investors have I bought thousands of nvidia chips To ensure their startups can get the strength they need. Andreessen Horowitz reportedly meta now owns).
But, without power, calculations are worthless. To ensure their data centers remain juiced, big tech companies have been rushing to buy Solar farm,Buy nuclear power plantand ink treatment Geothermal startups.
So far, Openai has been relatively quiet in this regard. CEO Sam Altman OKLO,,,,, spiraland Exowattbut the company itself isn’t putting money into the space like Google, Meta or Amazon.
With a 4.5 GW calculation protocol, it may change soon.
The company may play an indirect role, paying Oracle to handle physical infrastructure (it has extensive experience), just as Altman invests in startups that align with OpenAI’s future power demand. This will undoubtedly make the company’s Asset Light, which will undoubtedly please investors and align with other software-centric AI startups rather than with older technology companies that do not bear the burden of expensive infrastructure.
