Nebius raises capex guidance to $25B as AI infrastructure demand surges through 2027

Nebius Accelerates AI Infrastructure Investments Amid Soaring Demand

Nebius has dramatically increased its annual capital expenditure guidance to a range of $20 billion to $25 billion, up from its previous forecast of $16 billion to $20 billion, as the company races to meet surging demand for AI-related computing power 2

2

. The revised forecast signals the company's aggressive push to capture a larger slice of the rapidly expanding AI cloud expansion market, where demand for Nvidia GPUs and computing platforms continues to outpace supply.

The company reported quarterly capex of approximately $2.5 billion in the first quarter, a dramatic increase from $544 million a year earlier, slightly exceeding analyst estimates of $2.4 billion 1

1

. These investments in GPUs focused primarily on procuring graphics processing units and data center hardware to support the company's core AI infrastructure business. CEO Arkady Volozh highlighted the intensity of demand, noting that multiple customers were competing for every GPU brought online, providing the company with clear demand visibility through 2027 2

2

.

Strategic Acquisitions and Long-Term Contracts Drive Growth

Nebius has been expanding its footprint through strategic acquisitions and securing long-term contracts with major technology companies. Earlier this month, the company agreed to acquire AI startup Eigen AI for approximately $643 million to strengthen its inference platform and establish a stronger U.S. presence 1

1

. The acquisition represents a key move to enhance the company's AI capabilities and competitive positioning in the North American market.

The company also signed a significant deal with Meta to provide up to $27 billion worth of computing capacity over five years 1

1

. This partnership, alongside existing relationships with Microsoft, positions Nebius as a critical infrastructure provider for some of the world's largest technology companies. The company's client base demonstrates the growing reliance on specialized AI infrastructure providers to support increasingly complex machine learning workloads.

Revenue Growth Amid Margin Concerns

Despite heavy spending, Nebius delivered strong revenue growth in the first quarter. Revenue for the three months ended March jumped to $399 million, beating analyst estimates of $371.4 million 1

1

. However, analysts have flagged the company's heavy capital expenditure as a major concern, noting that aggressive expansion of its global data center footprint could pressure margins despite robust revenue growth 1

1

.

The company even paused its share buyback program in late 2024 to redirect capital toward expanding its core AI infrastructure business 1

1

. This decision underscores management's prioritization of growth over returning cash to shareholders in the near term. The concerns at Nebius mirror those at larger rival CoreWeave, which has projected between $30 billion and $35 billion in capital spending this year, warning that the ramp-up in investments could weigh on near-term margins 1

1

.

Expanding Physical Infrastructure to Meet AI Demands

Nebius announced the opening of a new site in Pennsylvania capable of eventually supporting 1.2 gigawatts of power capacity 2

2

. The massive power capacity reflects the energy-intensive nature of AI workloads and positions the company to handle increasingly demanding computational requirements. As AI models grow larger and more complex, access to substantial power infrastructure becomes a critical competitive advantage.

For investors and industry watchers, the key question centers on whether Nebius can maintain its growth trajectory while eventually improving profitability. The company's demand visibility through 2027 provides some reassurance, but the sustainability of current spending levels and their impact on long-term margins remains uncertain. With major technology companies continuing to invest heavily in AI capabilities, providers like Nebius that can deliver reliable computing capacity at scale stand to benefit, though the capital-intensive nature of the business creates execution risks that warrant close monitoring.