Khaberni - OpenAI signed a $38 billion agreement with AWS for cloud computing, a subsidiary of Amazon, to acquire additional development capabilities in artificial intelligence.
The startup developing "ChatGPT" continues its massive investments to acquire computing and storage power, whether from cloud computing service providers like AWS or chip manufacturers, to ensure its ability to compete in the AI race.
OpenAI aims to be the first company in this sector to develop a general model of artificial intelligence that fully mimics human intellectual capabilities.
Opinions differ on this point, as some experts expect the emergence of the general artificial intelligence model in the coming years, while others expect it to emerge within a decade at the earliest, or perhaps never at all.
OpenAI CEO Sam Altman stated on Monday that "developing advanced artificial intelligence requires massive and reliable computing resources."
He added, "Our partnership with +AWS+ enhances the extensive computing ecosystem that will support this new era and provide advanced artificial intelligence to everyone."
Last week, Altman revealed that OpenAI had secured $1.4 trillion in commitments from cloud computing service providers and the semiconductor sector.
These contracts will require 30 gigawatts of electricity, which represents more than 2% of the total installed electrical capacity in the United States by the end of 2023, according to figures from the U.S. Energy Information Administration (EIA).
Some investors are increasingly expressing their doubts about the huge deals by OpenAI, which expects its total revenue this year to be around $13 billion and which, according to Sam Altman himself, will not be profitable until 2029.
When asked about this issue in an episode of the "BG2 Pod" podcast that aired on Friday, Sam Altman expressed his discontent with the issue, saying that OpenAI will achieve "much higher" revenues than these estimates.
He added, "We expect our revenues to continue to grow strongly."
- Nvidia in a Leading Position -
The seven-year deal with Amazon Web Services (AWS) provides OpenAI with immediate additional resources in cloud computing, with full capabilities to be available by the end of 2026.
The cloud infrastructure to be provided by Amazon Web Services, based in California, will primarily rely on Nvidia's graphics processing units (GPUs), which are considered the most powerful in the market.
The California-based company announced that these processors will not only be used to work on new OpenAI models but also to operate "ChatGPT" and process requests from more than 800 million users weekly for this interface.
This news led to an increase in the Amazon stock price, which rose by approximately 4.46% at around 3:35 PM GMT in the New York Stock Exchange.
Nvidia stock also performed well (+2.55%), supported by the agreement between OpenAI and Amazon and by another announcement about Microsoft renting additional chips and servers from the cloud service provider "Iren" for a total value of $9.7 billion.
"Microsoft," the main partner of "OpenAI," which holds a 27% stake in it after investing more than $13 billion, has been open for months to the idea of the startup, based in San Francisco, seeking to provide cloud capacity from other companies.
The application submitted to "Iren" also shows that "Microsoft," despite being a cloud provider itself, is no longer able to meet all the data storage and processing needs of its clients.
