Back to News
IndustryAI Understanding briefing

Nebius secures 50 MW of AI data‑center capacity from AIB in a 12‑year deal

Nebius, an AI infrastructure provider, signed a 12‑year lease for 50 MW of IT capacity at an AIB Data Centers site in the southeastern United States, addressing a key power‑constraint for AI workloads.

4 min readRead the linked source
Source-provided image accompanying Nebius secures 50 MW of AI data‑center capacity from AIB in a 12‑year deal
Source referenceSource recorded
Publisher
w.media
Source type
Linked source — primary-source status has not been established.
ContextUnderstand this in 60 seconds

Key terms

Inference
The runtime phase where a trained model generates predictions or outputs.
Compute
The processing resources required to train and run models, often measured in FLOPS or GPU hours.
Latency
The time between sending a request and receiving the model's output.

What happened

Nebius, an AI‑focused infrastructure provider, entered a 12‑year agreement with AIB Data Centers Ltd. to lease 50 MW of IT capacity at an AIB site in the southeastern United States. The capacity will be delivered from a two‑hall development that already has access to 65 MW of utility power under a 15‑year electricity supply contract, allowing the project to move forward without major new electrical upgrades. AIB will finance much of the build‑out through advance payments from Nebius, project‑level borrowing, and preferred equity. The deal adds dedicated AI‑training and capacity for Nebius’s cloud services and marks AIB’s first long‑term contract with a leading AI infrastructure operator after its recent Texas acquisition that brought its total contracted power to roughly 120 MW.

Nebius announced that it will lease 50 MW of IT capacity from AIB Data Centers Ltd. under a 12‑year contract. The capacity is part of a two‑hall development at an AIB site that already has a 65 MW utility power connection secured through a 15‑year electricity supply agreement. This existing power arrangement means the project can proceed without the need for major new electrical infrastructure upgrades, a key factor cited by Nebius’s Chief Product and Infrastructure Officer, Andrey Korolenko.

AIB plans to fund the development primarily through advance payments from Nebius, supplemented by project‑level borrowing and preferred equity. The company recently completed an acquisition in Texas that raised its total contracted power capacity to roughly 120 MW, and it is actively seeking additional long‑term agreements with AI and high‑performance‑computing operators. AIB’s CEO, Jerry Tang, described the Nebius contract as a “transformational milestone” that validates the firm’s strategy of securing power in attractive markets and converting it into stable, contracted revenue.

The agreement is intended to support Nebius’s AI cloud operations, providing dedicated capacity for both training and workloads. While the press release did not disclose pricing or specific service‑level terms, the long‑term nature of the lease suggests a stable cost structure for Nebius and a predictable revenue stream for AIB.

Source details: w.media

Why it matters

The agreement tackles a widely reported bottleneck in AI infrastructure: the availability of reliable, high‑capacity power. By securing a 50 MW block with an existing power supply, Nebius can bring new AI resources online faster than if it had to wait for new grid connections or build its own power plants. This accelerates the rollout of AI training and services for Nebius’s customers, potentially lowering and cost for large‑scale model development. For AIB, the contract validates its strategy of acquiring sites with abundant power and converting that into long‑term, contracted revenue streams, reducing reliance on equity financing and limiting shareholder dilution. The deal also signals growing demand for dedicated AI‑grade data‑center capacity in the southeastern U.S., a region traditionally less saturated with AI‑focused facilities compared with coastal hubs.

Power availability has emerged as a critical constraint for scaling AI workloads, especially for large language models that require sustained high‑energy consumption. By locking in a sizable block of power that is already grid‑connected, Nebius can avoid the lengthy permitting and construction timelines that typically accompany new data‑center builds, potentially shortening the time‑to‑market for new AI services.

For AIB, the contract demonstrates the viability of its business model: acquiring sites with abundant power and converting that into long‑term, contracted revenue. This reduces the need for additional equity financing, which can dilute existing shareholders, and provides a more predictable cash flow profile that may be attractive to investors and lenders.

The deal also highlights a geographic shift in AI infrastructure investment toward the southeastern United States, a region that offers lower land costs and, in some cases, more favorable power pricing compared with traditional coastal data‑center hubs. This could encourage other AI operators to explore similar locations, diversifying the regional AI ecosystem.

Interactive Mechanism

Interactive Mechanism: How It Actually Works

Explore the underlying technology behind this development interactively.

Agent Lifecycle Stage:
1
User Intent & Planning: "Audit customer refund request #4092 and settle payment."
2
Tool Calling: Emits structured JSON call crm_get_transaction(id='4092').
3
Guardrail & Verification:Paused: High-value action requires human operator sign-off.
4
Final Settlement: Refund recorded, email receipt dispatched, and audit log stored.
Core takeaway: An AI agent is not just a language model—it is a closed loop of planning, tool invocation, and environment feedback. Production systems require self-healing retries and strict human approval guardrails.
Interactive Concept Check+10 Points
Future of AI Quiz

What should a useful AI forecast state?

What to watch next

Watch for additional long‑term power‑capacity contracts that AIB may secure with other AI or high‑performance‑computing operators, as the company has indicated a broader push for such agreements. Nebius’s rollout timeline and the speed at which the 50 MW becomes operational will indicate whether the “time‑to‑power” constraint is truly mitigated. Industry observers should also monitor regional power‑grid policies and pricing, which could affect the economics of similar deals in the southeastern United States. Finally, any announcements of further financing rounds or equity moves by AIB could reveal how the company balances debt, preferred equity, and advance payments to fund its expansion.

Future contracts that AIB may sign with other AI or HPC operators, which would indicate whether the power‑first strategy can be scaled across multiple sites.

Nebius’s deployment schedule and the actual date when the 50 MW becomes operational, which will test the claim that “time‑to‑power” is the binding constraint today.

Regulatory or policy developments affecting power pricing and grid capacity in the southeastern U.S., as changes could impact the economics of similar agreements.

AIB’s financing activities, including any new preferred‑equity issuances or debt facilities, which could reveal how the company balances capital sources to fund its expansion.

Related guides & quizzes

Found this useful?