Key conclusions
- NVIDIA OpenAi transaction for $ 100 billion is unprecedented: This strengthens the role of the company not only as the manufacturer of the chips and the supplier of graphic processors, but also as an investor for shares and the basis of the economy of AI.
- Vision for the construction of infrastructure is huge: 10 gigavatts of computing power (millions of graphic processors) can provide enough electricity for millions of houses, but also causes significant environmental problems.
- The risks are growing: Deloitte, UNEP and EESI warn that data -oriented data centers can double the global demand for electricity by 2030, are delivered on intense water and destabilize an electric grid.
- NVIDIA consolidates power, but risks revaluation: The transaction provides Openai as a client and increases dominance in the field of US AI, but also exposes NVIDIA to the risk of justice, political control and excessive dependence on certain TSMC factors.
Nvidia announced a commitment up to $ 100 billion in investments In Openai, which makes him the largest transaction on the infrastructure of AI so far.
The agreement signals a significant shift: NVIDIA is not only a supplier of the most popular graphic processors in the world, but also a direct investor in one of its leading customers. In turn, NVIDIA will receive shares in Openai, which is even more connected with the judges of these two leaders of the industry.
This is a decision for the general director of NVIDIA Jensen Juang to accelerate the “industrial revolution of AI.” However, for investors, this raises the question: is it a fuel to advance AI further to the mainstream, or is it a risky overload that can strain energy nets and challenging the limits of financial stability?
Details of the transaction and justification
NVIDIA commitment in the amount of 100 billion US dollars will be fulfilled for several years, starting with the initial investment of $ 10 billion after the completion of the agreement.
In turn, NVIDIA will receive shares in Openai and play a key role in the deployment of at least 10 gigavatts of computing power of artificial intelligence, enough to own millions of households. It is expected that the first Gigavatt will enter the Internet in the second half of 2026, and additional power will appear on the Internet, as NVIDIA plans to expand.
The strategy reflects the past nvidia movements with CoreweaveIN IntelAnd XaiWhere he used his balance to ensure demand for his graphic processors and strengthen its ecosystem.
The Director General of OpenAI Sam Altman described Compute Power as the main “fuel” that stimulates the progress of AI and the growth of income. Jensen Juan from Nvidia added that the computational demand “passes through the roof”.
Obviously, investors share this mood: NVIDIA shares increased by almost 4% after the announcement, adding more than $ 120 billion to the market value – more than the entire first stage of investment.
Creating a super-infrastructure of AI
NVIDIA in the amount of 100 billion US dollars will finance one of the most ambitious projects for the AI infrastructure: deployment of 10 gigavatts of computational capabilities for GPU on GPU on Openai
Here is a simple way to understand this number: one gigavatt can bring 876,000 houses on averageThe field shows both the size and the pressure that such objects will put on energy nets.
Planning related to the network of new data processing centers filled with millions of NVIDIA graphic processors, including the upcoming Vera Rubin platform, optimized for training and output of the next generation. It is expected that the first gigavatt of throughput will appear in the second half of 2026 with a further expansion at later stages.
Cooling these data centers can use up to 40% of their powerAccording to Energy Consultance 174 Power Global, causing growing environmental problems.
Nevertheless, the project corresponds to a wider Stargate initiative: a supporting Trump plan to create mega centers that strengthen the US leadership in the AI infrastructure.
Risks: power, water, politics and money
Under the hood of the headline for $ 100 billion. The United States is a set of serious risks that can form both the future of AI and NVIDIA.
The most obvious and immediate problem is the demand for energy. Deloitte ratings that the global data centers will consume about 536 teravatt hours (TWH) electricity in 2025. However, by 2030, unprecedented AI growth can easily push this figure up to 1000 TVHh, almost doubleting the trail of the sector.
The cooling of these objects adds another voltage layer. A Program environment (UNEP) warned about the growing use of water to disperse heat from these data centers: especially unreliable anxiety for regions that are already faced with a water shortage.
At the same time, The Institute of Environmental and Energy Research warns That the thirsty authorities of Clasters can emphasize local nets, which makes large -scale deployments of a political point of view.
From a geopolitical point of view, these risks only intensify. NVIDIA Director General Jensen Juan recently admitted This is “you cannot overcome the magic that is TSMC”, emphasizing both NVIDIA and Openai, remain in the Taiwanese production of chips: a potential weak link in the relations of the United States and China.
Finally, Nvidia takes on the direct risk of justice. Accepting Openai shares in exchange for financing infrastructure, NVIDIA connects itself with a startup with uncertain financial stability. If Openai stumbles, NVIDIA may be risk of reducing, despite the fact that he corrected the boom, which he helped to light.
Can NVIDIA keep all this together?
The Openai transaction emphasizes NVIDIA ambitions as not just a graphic processor supplier: it positions the company as the basis of the AI economy.
On the other hand, the obligation for $ 100 billion covers the NVIDIA openly as a long -term client, stabilizing the key player of AI in Openai, agreeing with the goals of the United States to ensure technological leadership. It also strengthens the already dominant ecosystem of the GPU from NVIDIA, which makes it even more difficult for competitors.
But the risks are also clear. Having become a supplier, an investor and gatekeeper of the ecosystem, Nvidia risks excessive concentration. Meanwhile, Openai receives some so necessary stability, the account relying even more deeply on the equipment and capital of NVIDIA.
For AMD, Intel and other applicants, a competitive gap in Ecosystems GPU and AI can currently be expanded even further.
Conclusion: the NVIDIA rate can provide its graphic processors, and its influence remained central in the continuing boom of AI. Nevertheless, with such a huge power, increased political control and environmental rollback come. We see that both of these things are already playing out, and this will only intensify further.
The editorial policy of Tech Report is focused on providing useful, accurate content, which offers real value for our readers. We work only with experienced writers who have specific knowledge on topics that they cover, including the latest developments in the field of technology, confidentiality on the Internet, cryptocurrencies, software and much more. Our editorial policy guarantees that each topic is studied and supervised by our internal editors. We support strict journalistic standards, and each article is 100% written by real authors.
