NVIDIA and OpenAI have announced a strategic alliance of historical dimensions: the deployment of at least 10 gigawatts of NVIDIA systems for the next generation of OpenAi’s infrastructure, which will mean millions of GPUS dedicated to training and executing its future models. The magnitude of the agreement is evident in the figure that has been made known, and that is that the green giant will invest up to 100,000 million dollars in OpenAIas each gigavatio unfolds. The first phase will start in the second half of 2026 on the Nvidia Vera Rubin platform.
According to the official statement, this infrastructure of 10 gigawatts will be built in collaboration with different strategic partners and will be progressively deployed with the aim of covering the growing demand for training and use of large -scale models. Each power gigavatio translates into the installation of GPUS, so the agreement is also a massive engineering and logistics project. But Nvidia will not be limited to providing hardware: The deployment will include the entire technological battery, from refrigeration systems and high -speed networks to its optimized software for the management of AI work loads.
OpenAI, on the other hand, declares that this alliance will allow you to make a qualitative leap in the speed and efficiency of its developments, ensuring access to computer resources on a scale that few actors in the world can guarantee in the same scenario. It is a long -term commitment, which seeks to anticipate the needs of the next generations of foundational models and the expansion of the services that the company directs to companies and consumers, to date, as a reference of the sector.
This announcement comes after months of rumors and negotiations that already anticipated the possibility that now materializes. In fact, in last summer information about the entry of Apple and Nvidia were leaked into an OpenAi financing round, with an eye on strengthening their strategic ties according to the interests of each party. While Apple worked in the integration of Chatgpt in iOS and in its Apple Intelligence project, Nvidia negotiated to join an estimated investment around 100 million dollars, which would have meant its third major capital injection in the company after those of 2019 and 2023.
The moment in which it occurs This agreement is also very favorable for Nvidiacompany that has marked a historical quarter with revenues of 46.7 billion dollars, that is, 56% more than the previous year, promoted above all by the sale of accelerators for artificial intelligence. The Data Centers Division already represents 88% of its business, with more than 41,000 million invoiced. It is not surprising that with such figures, Nvidia confirms its domain in the AI race, to the point that its CEO, Jensen Huang, openly speaks of a new industrial revolution with forecasts of between 3 and 4 billion dollars in spending on infrastructure of AI here to 2030.
In summary, Nvidia thus consolidates his already dominant position and does it just a week after announcing another operation of enormous relevance, such as his investment of 5,000 million dollars in Intel. A figure that, although it is overshadowed by the magnitude of the agreement with Openai, makes Nvidia one of the main investors of the signing of Santa Clara. In this case, the Agreement contemplates that Intel Fabrique Socs X86 to be integrated into the Chiplets of the NVIDIA GPU RTX, with the objective of shaping hybrid systems that reinforce their competitiveness and, at the same time, underpin the financial recovery of Intel.
