As artificial intelligence (AI) drives unprecedented computing power and pushes data centre infrastructure to its limits, Chemours has unveiled a new liquid cooling technology.

 

This is aimed at tackling one of the industry's biggest challenges: managing soaring heat loads while reducing energy and water consumption.

 

The Faizeon portfolio of two-phase liquid cooling solutions for AI, high-performance computing (HPC) and high-density data centres includes dielectric fluids designed for both direct-to-chip and immersion cooling applications.

 

The technology uses phase change, whereby a liquid absorbs heat and turns into vapour, enabling more efficient heat removal from high-power processors and servers.

 

In independent third-party testing, a two-phase direct-to-chip system using Faizeon 200 fluid maintained a graphics processing unit (GPU) at temperatures up to 19 degrees Celsius lower than a conventional single-phase, water-based cooling fluid, while operating at a lower coolant flow rate.

 

The launch comes as rising AI workloads drive higher chip power consumption and rack densities, placing increasing pressure on conventional data centre cooling systems.

 

“AI’s rapid growth is increasing the demands on data center infrastructure, and cooling is becoming a critical constraint,” said Denise Dignam, President and CEO at Chemours.

 

“With our liquid cooling solutions, we are applying more than 90 years of thermal management expertise to help data centers manage higher heat loads while improving performance and dramatically reducing energy and water demand. Our Faizeon liquid cooling team is collaborating across the data center ecosystem, to bring together the expertise, testing, and support needed to confidently deploy two-phase liquid cooling at scale.”

 

The Faizeon portfolio comprises five fluids designed to address different cooling requirements and infrastructure configurations:

 

Faizeon 100: A low-pressure, two-phase direct-to-chip fluid designed for systems where operating pressure, materials compatibility and infrastructure requirements are key considerations.

 

Faizeon 110: A low-pressure direct-to-chip fluid with an intermediate boiling point and zero ozone depletion potential, developed for advanced cooling systems.

 

Faizeon 200: A high-performance direct-to-chip fluid and industry standard for medium-pressure systems, designed to balance performance and environmental considerations in high-density AI and HPC applications.

 

Faizeon 220: Chemours' highest-performance two-phase direct-to-chip fluid, targeting applications requiring rapid heat removal, including high-density GPU environments, advanced AI clusters and HPC deployments.

 

Faizeon 500: A two-phase immersion cooling fluid, formerly known as Opteon™ 2P50 dielectric fluid, designed for full-server immersion cooling. It offers dielectric performance, full-server heat capture, support for ultra-high-density AI deployments and potential for waste heat recovery.

Lower energy and water consumption in focus

 

Beyond thermal performance, Chemours is positioning the Faizeon range as a solution to the growing energy and water demands of data centre operations.

 

Chemours is also working with organisations across the data centre ecosystem to advance the development and deployment of two-phase liquid cooling technologies.

 

Its collaborations include engagements with the U.S. Department of Energy's Oak Ridge National Laboratory (ORNL), NTT DATA, 2CRSi, DataVolt, Munters, Accelsius and ZutaCore.

 

These partnerships are intended to bring together industry expertise, testing capabilities and technical support to accelerate the adoption of advanced cooling solutions.

 

With AI infrastructure expanding rapidly and conventional cooling approaches facing mounting pressure, the launch of Faizeon™ highlights the growing importance of advanced thermal management technologies in supporting the next generation of high-performance computing.