OpenAI And Nvidia Push AI Into Industrial Age

In the global battle to dominate artificial intelligence, computing power has become the new currency. Nvidia is set to inject up to 100 billion dollars into OpenAI to build one of the most ambitious AI infrastructures ever conceived. This partnership marks a turning point, as the time for laboratory promises is over, and the era of massive AI industrialization begins. Such an initiative could reshuffle the cards of the sector and redefine technological power relations on a global scale.

An AI infrastructure worth 100 billion : Nvidia accelerates the pace

While the United Kingdom relies on the American firm to build its AI infrastructure, Nvidia has just revealed it signed a letter of intent to build a colossal AI infrastructure in partnership with OpenAI, worth $100 billion. This historic project aims to deploy 10 gigawatts of computing capacity, a power equivalent to that consumed by several million households.

We're proud to announce a landmark partnership with @OpenAI to build new gigascale AI factories using millions of NVIDIA GPUs. 🤝

This partnership will supply 10 gigawatts of GPUs to fuel @OpenAI's data center growth. pic.twitter.com/CYEB2PdfWY

— NVIDIA (@nvidia) September 22, 2025

This partnership marks a new stage in Nvidia’s strategy, moving from technology provider to strategic investor. “The demand for computing power is exploding”, said Jensen Huang, the company’s CEO. “This partnership aims to build an infrastructure that allows AI to move out of laboratories and into the real world. It is the industrial revolution of AI that is coming”, he added.

Following this announcement, Nvidia (NVDA) shares jumped 4%, reaching $184.16, a sign of market enthusiasm for this unprecedented alliance.

The terms of the deal reveal progressive financial commitments and a planned ramp-up over several years. Key points to remember :

• The first tranche of 10 billion dollars will be invested upon signing the agreement ;

• Nvidia will receive shares in OpenAI, making it a direct player in AI development ;

• Data centers will be equipped with millions of Nvidia GPUs, including the Vera Rubin platform, designed for future AI models ;

• The first gigawatt tranche is expected to be operational by the second half of 2026 ;

• This project is part of the continuation of the Stargate program, also supported by Microsoft and Oracle, aimed at building a global infrastructure dedicated to generative AI ;

• Sam Altman, CEO of OpenAI, welcomed this breakthrough : “it’s the fuel we need to improve our models, generate revenue, everything”.

Thus, AI enters an industrial phase, where computing power becomes the real strategic lever. Nvidia, betting 100 billion on the future, no longer just sells chips but is building the technological backbone of tomorrow’s artificial intelligence.

A colossal energy and environmental cost

While the ramp-up of AI infrastructure excites investors, it also triggers growing ecological concerns. Deploying 10 gigawatts of energy, equivalent to the consumption of millions of households, poses major challenges in energy and sustainability.

According to 174 Power Global, the cooling system alone of data centers can absorb up to 40 % of their total electricity consumption. As power needs increase, the entire energy chain could be under pressure. This year, data centers are expected to account for about 2 % of global electricity consumption, according to Deloitte. This figure could double by 2030, exceeding 1,000 terawatt-hours, notably due to the accelerated expansion of AI infrastructures, as evidenced by recent Northern Data purchase by Rumble.

The impacts do not stop at electricity. The United Nations Environment Programme (UNEP) has warned about the massive use of water to cool these computing centers. Meanwhile, the Environmental and Energy Study Institute (EESI) warns that power grids are already heavily strained by this growing demand, a situation that could quickly become critical without regulation or innovation in energy management.

In the face of this reality, the industrial shift taken by OpenAI and Nvidia could well impose a new type of responsibility on the tech ecosystem: that of energy efficiency, as crucial as algorithmic performance.

In this frantic race to build artificial intelligence infrastructure, some observers recall that bitcoin, for its part, already operates on a decentralized network with unparalleled computing power. Powered by millions of machines worldwide, the hashrate of the bitcoin protocol today exceeds the majority of institutional supercomputers.

Can this AI revolution be sustainable in the long term? As models become more powerful, their energy cost explodes. If this partnership symbolizes the dawn of a new industrial age founded on a more human artificial intelligence as promised by Sam Altman, it also forces a rethink of the balance between innovation, environment, and critical infrastructures. The future of the sector will depend as much on its technological capabilities as on its ability to respond to these systemic challenges.