NVIDIA plans to provide OpenAI with up to USD 25 billion in financing guarantees, specifically supporting SoftBank’s development of a 10-gigawatt hyperscale data center in Ohio. This move dispels prior rumors that NVIDIA was scaling back its support for OpenAI. By shifting from equity investment in OpenAI to project financing guarantees, NVIDIA not only avoids valuation negotiations but also secures massive future compute capacity orders from OpenAI.
$NVIDIA (NVDA.US)$ It is currently in discussions with OpenAI regarding financing guarantees of up to USD 250 billion for a data center project.
On July 26, according to The Wall Street Journal, NVIDIA is negotiating to provide approximately USD 250 billion in guarantees to help OpenAI lease computing power from a massive data center project. The deal would enable OpenAI to lease a 10-gigawatt data center hub being developed by SoftBank Group in Ohio. SoftBank founder Masayoshi Son previously stated that the project could cost as much as approximately USD 500 billion.
This potential transaction completely dispels earlier market rumors that 'NVIDIA was scaling back its support for OpenAI by hundreds of billions of dollars.' Shifting from a shelved direct investment of USD 100 billion—paused due to OpenAI’s IPO preparations—to a USD 250 billion project financing guarantee, NVIDIA has not only significantly increased the scale of its financial commitment but also upgraded the nature of its support.
Bypassing equity-related negotiations, the USD 250 billion guarantee reshapes the capital alignment model
Previously, there were market reports suggesting that NVIDIA had ruled out providing hundreds of billions of dollars in financial support to OpenAI. According to earlier plans, NVIDIA had considered a USD 100 billion investment proposal, which was put on hold due to OpenAI’s IPO process and subsequently adjusted to the current round of USD 30 billion in investment.
Now, the emergence of the USD 250 billion guarantee plan indicates that NVIDIA has not scaled back its financial support for OpenAI but has instead fundamentally changed its approach. By shifting from direct equity investment to project financing guarantees, NVIDIA has skillfully avoided valuation disputes and equity dilution risks associated with OpenAI’s pre-IPO phase. This model not only leverages substantial infrastructure projects with relatively low capital outlay but also ensures that OpenAI’s future massive computing power procurement orders continue flowing to NVIDIA, maximizing both commercial returns and ecosystem control.
Anchoring supercomputing hubs and government-controlled power to accelerate the realization of the 'trillion-token factory'
The core of these negotiations is that the guaranteed funds will be strictly earmarked for OpenAI’s hyperscale data center hub project, aimed at building one of the world’s largest AI computing centers. Notably, the project also involves electricity resources controlled by the U.S. government, signaling an expansion in the dimensions of AI infrastructure and a clear indication of public-private collaboration.
This aligns closely with NVIDIA CEO Jensen Huang’s recent vision of a 'trillion-token factory.' Huang explicitly stated that the focus of AI development is shifting from model training to inference, and that data centers will evolve into AI engines processing unstructured data. To realize this vision, OpenAI needs to build unprecedentedly large data center hubs, yet such capital-intensive infrastructure faces significant funding and energy constraints. By providing massive guarantees and facilitating access to government-controlled power resources, NVIDIA is effectively removing both financing and energy barriers to OpenAI’s computing capacity expansion.
Financial exposure challenges and far-reaching implications for the AI infrastructure landscape
Although NVIDIA currently enjoys strong cash flow, the scale of the $250 billion guarantee remains substantial and will inevitably have a profound impact on its financial statements. As the guarantor, NVIDIA will assume significant contingent liability risk. Should OpenAI’s data center projects fail to generate expected returns or experience debt defaults, NVIDIA’s financial risk exposure would be forced to expand. Therefore, the final guarantee terms, risk-sharing mechanisms, and regulatory approvals—along with their feasibility—remain central points of negotiation at this stage.
From an industry structure perspective, this move will further solidify NVIDIA’s moat in the AI computing power domain. Coupled with NVIDIA’s recently disclosed $95.2 billion commitment to supplier purchases and approximately $90 billion in ecosystem investments over the past 16 months—spanning more than 145 companies across the AI value chain—NVIDIA is seeking to establish absolute dominance in next-generation AI infrastructure through multiple barriers: financial instruments, hardware capacity, and energy access. This not only secures long-term demand from its largest customer but also significantly raises the $Advanced Micro Devices (AMD.US)$ 、 $Intel (INTC.US)$ entry barriers for competitors such as
Editor/KOKO