Meta and Nebius: a $27 Billion Partnership for AI
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisis ton rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
Meta Partners with Nebius to Secure $27 Billion in AI Power
Meta has just struck a colossal deal with Nebius to bolster its infrastructure dedicated to artificial intelligence. The total value of the partnership reaches $27 billion.
As part of this agreement, Nebius will provide Meta with dedicated computing capacity estimated at $12 billion. This infrastructure will be deployed across multiple sites and is expected to be available starting early next year.
The announcement was made jointly by the two companies on Monday, March 16. The project will also include one of the first large-scale deployments of the NVIDIA Vera Rubin platform developed by NVIDIA.
In exchange, Meta has committed to purchasing up to an additional $15 billion in computing capacity over a five-year period, provided that these resources are not sold to other clients.
This partnership is part of a series of agreements already concluded by Nebius. In November, the company signed an initial $3 billion contract with Meta. It also announced a $17.4 billion agreement with Microsoft in September. Last week, the company even finalized a $2 billion deal with Nvidia.
More Information on Nebius
Based in Amsterdam, Nebius emerged from the restructuring of the former Russian internet group Yandex. After divesting its historic search engine, the firm repositioned itself in cloud infrastructure services aimed at AI projects.
Nebius belongs to a new generation of companies. Unlike traditional hyperscalers, these players focus almost exclusively on providing optimized GPU power.
Nebius's infrastructure relies on recent NVIDIA GPUs and high-speed networks to support these computing loads. They are explicitly designed for training, fine-tuning, and inference of large-scale AI models.
The company recently announced its intention to acquire Tavily, a provider of agent-based search technologies. The goal is to combine access to real-time data with Nebius's AI platform, especially as more companies adopt AI systems capable of acting autonomously.
Financially, the group's shares listed on Nasdaq, which closed at $112.50 on Friday, have risen by 35% since the beginning of the year. This growth now brings the startup's market capitalization to approximately $28.6 billion.
However, the rapidly expanding company remains unprofitable. In February, Nebius reported a net loss of $250 million in the fourth quarter, with revenues of $228 million.
Nonetheless, the company anticipates rapid growth. It estimates that its revenue could reach an annualized rate between $7 and $9 billion by the end of this year, compared to $1.25 billion by the end of 2025.
Another Deal for Meta
This agreement is part of Meta's massive spending strategy in the field of artificial intelligence. The group is ramping up investments, particularly through the acquisition of startups like Manus and Moltbook.
At the same time, the company is also strengthening its teams by offering multi-million dollar salaries to attract top AI specialists.
According to Reuters, Meta also plans to invest between $115 and $135 billion this year to develop its data centers, chips, and all the necessary infrastructure to train increasingly powerful AI models.
With such a level of spending, Meta now joins the ranks of the world's largest investors in AI infrastructure, alongside Google, Microsoft, and Amazon.
The difference, however, is that these companies have public cloud operations that allow them to rent this computing power directly to their clients. Meta, on the other hand, primarily uses this infrastructure for its own AI models. The group does not have the same monetization instrument through cloud services.
Additionally, a recent article published by Reuters claimed that Meta is preparing for a significant wave of layoffs. This could affect up to 20% of its workforce to offset the scale of its investments in artificial intelligence. However, the company has denied this information through a spokesperson.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.