Meta Spends Another $27B on AI Infrastructure With Nebius
The agreement, one of the biggest AI compute deals ever, arrives even as AI vendors, including Meta, are reportedly considering mass layoffs.
Cheng Xin via Getty Images
The Facebook and Instagram parent company on Monday agreed to a new multi-billion-dollar deal with AI infrastructure supplier Nebius.
Under the terms of the five-year agreement, revealed on Monday, the Amsterdam-based neocloud vendor will supply the social media giant $12 billion of dedicated capacity across a number of locations.
This will include what Nebius said will be one of the first large-scale deployments of Nvidia’s Vera Rubin platform, with delivery set to start next year.
In addition, Meta has committed to buying available compute capacity across specific Nebius clusters over the next five years, to a total value of $15 billion.
“We are pleased to expand our significant partnership with Meta as part of securing more large, long-term capacity contracts to accelerate the build-out and growth of our core AI cloud business," Nebius CEO Arkazy Volozh said in a release.
Nebius stock soared 15% in midday trading on the news.
Related:Nebius to Build One of Europe’s Largest AI Factories in Finland
The deal came less than a week after AI chip giant Nvidia revealed that it was investing $2 billion in Nebius.
The agreement follows a previous deal Nebius brokered with Meta last November, which was valued at $3 billion. That development came quickly after a separate deal worth up to $19.4 billion for Nebius to deliver GPU capacity to Microsoft for a new data center in Vineland, New Jersey.
For Meta, the investment is another major outlay, as the fierce race to secure AI infrastructure shows no sign of slowing.
In late February, Meta revealed a major deal with Nvidia to use the chip giant’s Grace central processing units, as well as Blackwell and next-generation Vera Rubin GPUs, in its data centers.
The financial terms of the deal were not disclosed, but were said to extend to tens of billions of dollars. CEO Mark Zuckerberg had previously suggested that Meta would spend up to $135 billion on AI in 2026.
Amid this dramatic level of investment, it is perhaps no surprise that rumors are now gaining traction that significant job cuts are on the horizon. Three separate sources told Reuters that sweeping layoffs could see about 20%of Meta’s workforce lose their jobs -- amounting to about 16,000 employees.
Meta has not commented on the speculation.
About the Author
Contributing Writer
Graham Hope has worked in automotive journalism in the U.K. for 26 years, including spells as editor of leading consumer news website and weekly Auto Express and respected buying guide CarBuyer.
Sign in to highlight and annotate this article

Conversation starters
Daily AI Digest
Get the top 5 AI stories delivered to your inbox every morning.
More about
reportBeyond Static RAG: Using 1958 Biochemistry to Beat Multi-Hop Retrieval by 14%
<p>Standard Retrieval-Augmented Generation (RAG) often falls short on complex, multi-hop questions because it relies on static "lock and key" query matching. If the information needed to answer a query is semantically distant from the original text, standard vector search simply won't find it.</p> <p>We've developed Induced-Fit Retrieval (IFR), a dynamic graph traversal approach that mutates the query vector at every step to discover semantically distant but logically connected information.</p> <p>The Core Results<br> We ran our prototype through a rigorous test suite of 30 queries across multiple graph sizes, up to 5.2 million atoms.</p> <p>14.3% higher nDCG@10 compared to a competitive RAG-rerank baseline.</p> <p>15% Multi-hop Hit@20 in scenarios where traditional RAG methods scored 0%.<
Meta will fund seven new natural gas plants to power its biggest data center yet
Meta will essentially foot the power bill for the $27 billion mega data center it's building in Louisiana. On Friday, the Wall Street Journal reported that the company struck a deal to fund the energy infrastructure needed for the project. Through a deal with Entergy Louisiana, Meta will fund seven new natural gas power plants, 240 miles of transmission lines and battery energy storage at three locations. The gas plants will have a combined power output of 5,200 megawatts, and the transmission lines will operate at 500 kilovolts. In addition, the company will help fund up to 2,500 MW of new renewable resources. It also agreed to a memorandum of understanding for future nuclear power development. The 4-million-square-foot Richland Parish, LA, data center will be Meta's biggest yet . It's cu

CausalPulse: An Industrial-Grade Neurosymbolic Multi-Agent Copilot for Causal Diagnostics in Smart Manufacturing
arXiv:2603.29755v1 Announce Type: new Abstract: Modern manufacturing environments demand real-time, trustworthy, and interpretable root-cause insights to sustain productivity and quality. Traditional analytics pipelines often treat anomaly detection, causal inference, and root-cause analysis as isolated stages, limiting scalability and explainability. In this work, we present CausalPulse, an industry-grade multi-agent copilot that automates causal diagnostics in smart manufacturing. It unifies anomaly detection, causal discovery, and reasoning through a neurosymbolic architecture built on standardized agentic protocols. CausalPulse is being deployed in a Robert Bosch manufacturing plant, integrating seamlessly with existing monitoring workflows and supporting real-time operation at product
Knowledge Map
Connected Articles — Knowledge Graph
This article is connected to other articles through shared AI topics and tags.
More in Analyst News
Caltech’s massive 6,100-qubit array brings the quantum future closer
Caltech scientists have built a record-breaking array of 6,100 neutral-atom qubits, a critical step toward powerful error-corrected quantum computers. The qubits maintained long-lasting superposition and exceptional accuracy, even while being moved within the array. This balance of scale and stability points toward the next milestone: linking qubits through entanglement to unlock true quantum computation.
This tiny chip could change the future of quantum computing
A new microchip-sized device could dramatically accelerate the future of quantum computing. It controls laser frequencies with extreme precision while using far less power than today’s bulky systems. Crucially, it’s made with standard chip manufacturing, meaning it can be mass-produced instead of custom-built. This opens the door to quantum machines far larger and more powerful than anything possible today.
New prediction breakthrough delivers results shockingly close to reality
Researchers have created a prediction method that comes startlingly close to real-world results. It works by aiming for strong alignment with actual values rather than simply reducing mistakes. Tests on medical and health data showed it often outperforms classic approaches. The discovery could reshape how scientists make reliable forecasts.
Seeing beyond the scan in neuroimaging
In this episode, we explore the intersection of AI, machine learning, and healthcare through the lens of neuroimaging and epilepsy diagnosis. Dr. Gavin Winston shares insights from his work using MRI data and machine learning to uncover subtle abnormalities in brain function. We discuss the cultural and ethical barriers to AI adoption in medicine, how predictive data analysis could transform the diagnostic workflow, and what the future holds for medical imaging in a world increasingly shaped by intelligent systems. Featuring: Gavin Winston – LinkedIn , Website Chris Benson – Website , GitHub , LinkedIn , X Daniel Whitenack – Website , GitHub , X Links: Detection of Epileptogenic Focal Cortical Dysplasia Using Graph Neural Networks: A MELD Study Machine Learning in Neuroimaging across Disci

Discussion
Sign in to join the discussion
No comments yet — be the first to share your thoughts!