The two companies aims to build a coherent and integrated ecosystem which ensures them total domination in the future AI market.
To those who thought that Nvidia would take a break after his investment frenzy in recent weeks, the international champion of AI fleas has just brought a scathing denial. 100 billion: this is the gigantic sum that the company will invest in Openai. It is neither more nor less of the biggest investment ever made in a private company, spraying the sums previously invested by Microsoft ($ 13 billion) and SoftBank (40 billion) in Openai. It will allow Sam Altman’s company to accelerate the construction of data centers for AI, which it has an insatiable need, for a combined power of 10 gigawatts (GW). Either the equivalent of the annual electricity consumption of a small country in Latin America, such as Costa Rica or Uruguay, or ten nuclear reactors. An investment that will allow Openai to buy fleas … nvidia
These data centers will naturally be equipped with the new generation of Vera Rubin chips, which will replace Blackwell technology from the second half of 2026. Suffice to say that Nvidia should quickly recover a good part of its investment. As much as the boundless confidence that investors place in Openai, despite the fears of an AI bubble, this agreement testifies to the formidable financial power that Nvidia conquered. Jensen Huang’s company has become a real cash machine thanks to its unrelated domination over the ia chips, that neither the threat of an energy shortage nor the attacks of China seem to be able to start. Several additional lessons can however be drawn from this spectacular investment.
An agreement that surfs the AI bubble
As we have seen: if Nvidia invests in a gargantuan sum in Openai, she will recover a good part via the flea orders that the chatgpt mother-in-law will spend him to equip its future brand new data centers. One can therefore wonder about the interest of an agreement which consists in transferring to Openai money which will then return to the pocket of Nvidia.
It would be to forget that the two companies have everything to gain that the perception of a race for AI that accelerates is strengthening. The valuation of Openai, currently estimated at $ 500 billion (unheard of for a private company) is notably based on the perception that AI will radically transform society, and that Optai is one of the best placed to derive the profits. A giant investment followed by the announcement of a frantic construction of overpowered data centers can only strengthen this perception.
Nvidia, for its part, has every interest in the race for AI continues to accelerate, since it leads companies like Openai, Microsoft, Meta, Google, Anthropic or Mistral to equip themselves in Nvidia chips, in a arms race where everyone fears to be overtaken by competition. It is therefore likely that this new agreement with Openai convinces it of it to review their investments in Nvidia flea.
OPENAI and NVIDIA forge a coherent AI ecosystem
This announcement also shows how the two giants of the AI intend to forge a coherent ecosystem which includes the entire AI value chain. In recent months, Openai has invested $ 300 billion to acquire computer power from Oracle and started to work on its own fleas with Broadcom.
She has also just announced the creation of five Giant AI data centers with Oracle and SoftBank as part of the Stargate project. All this denotes a desire to go up the value chain to go beyond the design of the models and master each step, from the creation of the chips to the construction and the operation of the data centers. Nvidia does the same. In the space of a week, the IA flea giant has invested $ 5 billion in Intel, the only American company with foundry capacities, as well as two billion sterling pounds in the young British shoots of AI, still testifying to a desire to master more and more bricks.
Dimitri Zabelin, a specialist in AI at Pitchbook, also commented on the agreement with Openai by saying that he was showing how Nvidia “consolidates his control over the AI value chain and strengthens his essential supplier position”. Indeed, to dominate the future of AI, excel in a specific technological brick will be insufficient: it will be necessary to offer a powerful and coherent ecosystem.
This is the thesis defended by Jacobid, Brusoni and Candelon researchers in their study called “The Evolutionary Dynamics of the Artificial Intelligence Ecosystem”, published in the journal Strategy Science. They analyze the way in which the AI market is dominated by a few “superstar” companies that have been able to set up complete and interconnected ecosystems, and highlight the role of the integration of the various technological stacks for anyone who wishes to strengthen its monopoly on AI. According to them, “the world of AI is characterized by the domination of a small number of large technological companies, whose use of AI (for example, in research, payments, social networks) has supported a large part of the recent progress, and which also ensure the supplies upstream of the necessary computing power (Cloud and Edge). These companies dominate the main academic institutions in AI research, their position. ”
What applies to cloud giants also applies to Nvidia. Beyond the power of its graphic processors, for the time being unrivaled, it draws for example its strength from its ecosystem, which society has patiently developed for twenty years already. An ecosystem which includes communities of developers who help each other and design bookstores; Companies that cause large models of AI and smaller models; And size clients like GAFAMs that set up the software bricks necessary to deploy NVIDIA GPUS in their data centers. She now seeks to go further by also being present in AI models, through her investment in Openai, and production (via Intel).
Openai’s judicious bet on Nvidia
The authors of the study cited above also believe that the latest changes in the AI market around inference and computer science on the outskirts promote, in this dominant ecosystem race, actors like Nvidia to the detriment of historic cloud giants. “As IT on the outskirts-that is to say made on smaller local devices, such as cameras and phones-is generalized, flea manufacturers such as Nvidia and manufacturers of devices like Huawei can enhance their devices compatible with AI and their new flea architectures, which make the Edge functions more efficient.”
By approaching Nvidia as she moves away from Microsoft, her historic privileged partner, Openai would make a winning bet on the future …
