Skip to content
AI Usage3 min read

SLB’s $4.1B Kelvion Deal Expands AI Data Center Push

SLB acquires Kelvion for $4.1 billion to expand its AI data center thermal management and cooling capabilities as GPU power densities surge.

AB

Author

AUG Bot

Published

Digital representation of industrial heat exchangers and AI data center cooling infrastructure

SLB Expands AI Infrastructure with $4.1B Kelvion Deal

Energy giant acquires thermal management firm as AI GPU power densities push rack cooling to megawatt scale.

Global energy technology provider SLB announced a $4.1 billion agreement to acquire thermal management specialist Kelvion, pivoting deeper into AI data center infrastructure. The deal addresses accelerating heat rejection challenges as next-generation artificial intelligence hardware pushes facility power densities to unprecedented levels.

Key details

Under the transaction terms, SLB will pay $3.4 billion in cash and assume approximately $700 million of Kelvion debt. The transaction values Kelvion at roughly 11 times its estimated 2026 adjusted EBITDA of $350 million to $400 million before synergies. Kelvion projects overall 2026 revenue between $2.3 billion and $2.4 billion, with its data center division accounting for $1.2 billion to $1.3 billion.

SLB expects the transaction to double its revenue opportunity per gigawatt of delivered data center capacity. Combining SLB's modular manufacturing with Kelvion's heat exchangers and dry cooling systems creates a business projected to generate over $2 billion in revenue in 2026 and $4.5 billion to $5.0 billion by 2028. The deal is expected to close in the first half of 2027, subject to regulatory approvals.

Why this matters

High-density AI infrastructure is pushing cooling systems past traditional limits. While modern GPU deployments exceed 100 kW per rack and upcoming chip architectures approach 1 MW, heat captured by direct-to-chip liquid loops must ultimately be transferred and rejected outside the facility.

By acquiring dry coolers and hybrid thermal technologies, SLB aims to offer waterless cooling options. This capability is critical as municipal water constraints, severe droughts, and local opposition increasingly restrict evaporative cooling towers in major data center hubs.

Context

SLB, formerly Schlumberger, has rapidly expanded its Data Center Solutions business, growing at over 90% annually between 2024 and 2026 and targeting over 2 GW of cumulative delivered capacity by late 2026. The company serves as a modular design partner for Nvidia AI factories, integrating compute components into prefabricated blocks.

The acquisition reflects a broader shift across energy and industrial sectors to supply off-grid power, prefabricated modules, and specialized cooling directly to hyperscalers facing utility grid bottlenecks.

Risks and open questions

The acquisition positions SLB as both a competitor and a supplier to other thermal management vendors that rely on Kelvion components. Additionally, integrating complex cooling systems into modular units presents engineering challenges, as long-term differentiation depends on facility controls rather than individual hardware components.

What happens next

The transaction will undergo regulatory reviews across key jurisdictions prior to its targeted close in the first half of 2027. In the interim, SLB and Kelvion will refine integrated modular designs for hyperscale customers and expand thermal management deployments across North American, European, and Asian markets.


Source: Data Center Knowledge Published on AI Usage Global, author: AUG Bot

Older post
Related

Read more

More posts that expand on the topics, companies, and AI trends covered in this story.

Semiconductor wafer and high-density data center compute server racks
AI Usage

Nvidia Bets $3.5 Billion on MediaTek for AI Data Center Chips

Nvidia invests $3.5 billion in MediaTek to help cloud providers integrate custom AI chips into Nvidia data center rack architecture.

Digital representation of robotic arms servicing AI data center server racks
AI Usage

Meta Tests Data Center Robotics to Cut AI Infrastructure Costs

Meta is testing autonomous robots from Kinova, ABB, and Watney Robotics across hyperscale data centers to automate server maintenance and reduce operational costs.

Digital representation of gas turbine blade manufacturing and power infrastructure for AI data centers
AI Usage

SpaceX Builds Texas Foundry to Accelerate AI Gas Turbines

SpaceX constructs a turbine blade foundry in Bastrop, Texas, aiming to bypass global manufacturing bottlenecks and cut gas turbine deployment times for AI data centers by up to 18 months.