Both Amazon’s $50 billion and Nvidia’s $30 billion investments are bound by enormous customer commitments from OpenAI and its partners.
Of Amazon’s investments, $35 billion will only be paid out “if certain conditions are met.” From what we’ve gathered, these terms also include rent from two gigawatts of Amazon’s Trainium AI accelerators and the deployment of OpenAI models and services in AWS. The cloud provider will also be the “exclusive third-party cloud distribution provider for OpenAI Frontier.”
Earlier this month it was announced that Frontier is the new version of OpenAI real estate agent builder aimed at business customers. Curiously, OpenAI was quick to emphasize that “nothing about today’s announcements in any way changes the terms of the relationship between Microsoft and OpenAI,” and that Azure remains the exclusive cloud provider of its stateless OpenAI APIs and first party products, such as Frontier.
Nvidia’s $30 billion investment appears to have similar provisions. In one blog post Published on Friday, the AI flagbearer announced an expanded partnership with Nvidia that would see the deployment of three gigawatts of inference and two gigawatts of training capacity built on the GPU crank’s Vera Rubin systems. The racks, which were announced at CES in January, are expected to ship in the second half of 2026.
The economics of data centers can be difficult to assess because they vary by region, but with an energy efficiency of 1.1, one gigawatt is sufficient for approximately 3,600 Vera Rubin NVL72 systems, assuming 250 kW per rack.
At one estimated cost of $8.4 million each, which is roughly $30 billion per gigawatt. However, compute is only responsible for about half of the cost of setting up a modern AI data center.
The costs of land, forage, energy and plumbing make up the rest. This brings the cost of building and deploying 5 gigawatt Vera Rubin accelerators to more than $300 billion. OpenAI is unlikely to take on this on its own. Instead, we expect CEO Sam Altman to broker purchasing deals through his hyperscaler and neocloud partners, as he has done with Oracle and Crusoe for its Stargate facility in Abilene, Texas.
Amazon is one of those partners. Although OpenAI plays into Trainium’s hands in today’s revelations, it does existing $38 billion partnership with Amazon, already announced in Novemberwas specific to Nvidia GB200 and GB300 systems. That contract has now been expanded to $100 billion over the next eight years.
Both Amazon and Nvidia’s investments are structured to guarantee a return on every dollar invested in OpenAI. The funding is essentially a discount on computing infrastructure that doesn’t dilute their revenues while boosting OpenAI’s valuation.
This kind of circular action has become commonplace during the rise of AI, and all the big boys are participating. In October, Nvidia competed with AMD issued OpenAI has been granted a warrant for approximately 10 percent of its shares. To claim this, Altman only needs to deploy six gigawatts of the chip designer’s Instinct accelerators. This week AMD copied and pasted that deal, which extends the same offering to social media and potential AI mogul Meta.
While Nvidia and Amazon are clearly engaged in financial engineering, the same does not appear to be the case for SoftBank. Notorious spendthrift Masayoshi Son plans to do just that invest another $30 billion into the company to keep the lights on as Altman and his team continue their quest for artificial general intelligence. SoftBank’s contribution will be paid out in three tranches of $10 billion, starting in April and ending in October.
Even with reportedly recurring revenue on an annual basis exceed With $20 billion and more than 50 million paying subscribers to its books, OpenAI will be dependent on external funding for a while. The company is not expected to be profitable until 2029. ®
#Amazon #Nvidia #open #wallets #capture #OpenAI #orders