
"OpenAI's explosive growth is causing a wide range of technology partners to accumulate unprecedented levels of debt. According to an analysis by the Financial Times, parties such as SoftBank, Oracle, and CoreWeave have collectively borrowed at least $30 billion. When the financing from investors and infrastructure suppliers is also taken into account, the total debt associated with OpenAI is now approaching $100 billion."
"Investors such as Blue Owl Capital and compute infrastructure companies such as Crusoe are heavily dependent on their collaboration with the AI start-up. Together, they hold around $28 billion in loans linked to computing power, data centers, or long-term hardware lease arrangements. In addition, a group of banks is negotiating new financing of approximately $38 billion for Oracle and data center developer Vantage. This loan should enable OpenAI to establish new locations needed to train future models on a large scale."
"According to the Financial Times, the internal view is that growth can best be facilitated when other parties make their balance sheets available for the necessary infrastructure. At the same time, the company has signed $1.4 trillion in multi-year commitments for chips and computing power this year, which stands in stark contrast to its expected revenue of approximately $20 billion."
Technology partners have borrowed at least $30 billion to support OpenAI's expansion. Including financing from investors and infrastructure suppliers, total debt linked to OpenAI is approaching $100 billion. Investors such as Blue Owl Capital and compute providers such as Crusoe hold about $28 billion in loans tied to compute power, datacenters, or long-term hardware leases. Banks are negotiating roughly $38 billion in new financing for Oracle and datacenter developer Vantage to build locations for large-scale model training. Most short-term financial responsibility rests with partners and their lenders while OpenAI plans to take on more debt over time. The company has signed $1.4 trillion in multi-year chip and compute commitments despite expected revenue near $20 billion, and it remains highly valued while scaling amid hardware shortages and rising energy demands.
Read at Techzine Global
Unable to calculate read time
Collection
[
|
...
]