DCX Liquid Cooling Systems has reported a 600% year-on-year increase in revenue in 2025, which it attributes to rising demand for liquid cooling from hyperscale operators, AI infrastructure providers and enterprises shifting away from air-based approaches.

It’s been no secret that liquid cooling is gaining in popularity as meeting the needs of AI workloads continues to dominate the conversation – but now we have concrete data on just how popular the systems are proving. DCX, which is based in Poland, had its best year ever in 2025 – with the 600% revenue jump just the icing on the cake for the firm. 

That’s because 2025 also saw the firm deliver a 5MW facility distribution unit (FDU), launch a new AI liquid cooling system reference design, and add three new hyperscale AI deployments to its project portfolio. DCX also said it expanded its workforce to more than 120 employees during the year.

600% revenue growth and hiring surge

DCX said its 600% year-on-year revenue growth was driven by ‘strong demand from hyperscale operators, global AI infrastructure providers, and enterprise customers transitioning away from air-based cooling’.

“This growth is not the result of a single breakthrough, but of years of disciplined execution,” said Tomasz Buk, Chief Executive Officer of DCX Liquid Cooling Systems. 

“Customers are choosing DCX because we combine engineering depth with delivery certainty. In a market moving this fast, trust and performance matter more than ever.”

To support an expanding pipeline of deployments, DCX increased its workforce more than fourfold in 2025, growing to over 120 employees worldwide. The company said the recruitment push focused on engineering, project management, commissioning, manufacturing and after-sales support.

5MW facility distribution unit delivered

To support further growth, DCX delivered what it’s calling “one of the most powerful Facility Distribution Units (FDUs) currently in operation”, capable of handling up to 5MW of cooling capacity within a single system.

According to the company, the unit was engineered to support high-density racks used in AI training and inference workloads, while maintaining availability, redundancy and precise thermal control. DCX also said the system was designed to reduce the cooling infrastructure footprint to help preserve revenue-generating IT space in hyperscale and AI-focused data centres.

“At this scale, liquid cooling is no longer just about removing heat, it’s about system architecture now,” added Maciek Szadkowski, Chief Technology Officer at DCX Liquid Cooling Systems. 

“Delivering a 5 MW FDU required rethinking redundancy, hydraulics, and control logic from the ground up. This project reflects how far liquid cooling has evolved and where it must go to support AI at scale.”

New reference design aimed at hyperscale densities

As the company increasingly receives interest in its solutions from hyperscalers, the firm introduced an AI liquid cooling system reference design, which it said was built to address limitations in traditional data centre cooling approaches.

The company said its reference design is intended to support rack power levels “ranging from 115 kW to hundreds of kW per rack”, and to improve efficiency through raised loop temperatures. It also said the design separates power and compute infrastructure from mechanical systems such as pumping and cooling loops, with the aim of improving flexibility and scalability as AI workloads evolve.

Hyperscale AI deployments and 500MW under contract

While the switch to liquid cooling across the board has seen DCX increase its revenues to record levels, the firm’s growing pipeline of customers should set it up for success in the future. In 2025, the firm secured and executed three hyperscale liquid cooling deployments for customers it described as being among the ‘world’s top ten AI companies’. While the firm did not name the customers, it said the projects involved ‘highly customised cooling architectures’ designed for extreme power densities, rapid scalability and uninterrupted operation.

Across all projects in 2025, DCX said it shipped and secured contracts for more than 500MW of liquid cooling capacity. It said this included large hyperscale deployments, AI-specific facilities and modular data centre solutions.

In addition, the company expanded its geographic footprint through deployments supporting 300MW-class data centres under construction in both North America and Europe.

Outlook for 2026

DCX believes that ‘significant growth’ will continue in 2026, citing demand and negotiated contracts, and said it plans to continue increasing system capacities and advancing its liquid cooling platforms.

“The industry is only at the beginning of this transition” added Maciek Szadkowski. 

“What we’ve delivered in 2025 sets the foundation for what comes next, and the next generation of AI infrastructure will depend on it.”