New Faizeon™ portfolio of two-phase direct-to-chip and immersion cooling fluids address rising AI heat loads with superior performance and reduced energy and water use
The Chemours Company has launched Faizeon™, an advanced portfolio of two-phase liquid cooling solutions tailored to resolve escalating thermal constraints in artificial intelligence (AI), high-performance computing (HPC), and high-density data center environments. Leveraging liquid-to-vapor phase-change physics, the dielectric fluids absorb and dissipate extreme heat loads from high-power processors and servers while substantially lowering cooling-related power consumption and nearly eliminating facility water use. The new portfolio comprises specialized direct-to-chip formulations including Faizeon™ 100, 110, 200, and the high-performance 220 fluid alongside Faizeon™ 500, a dielectric immersion cooling fluid formerly known as Opteon™ 2P50. In independent evaluations, a two-phase direct-to-chip configuration utilizing Faizeon™ 200 kept GPU temperatures up to 19 degrees Celsius lower than conventional single-phase water systems despite requiring lower coolant flow rates.
Also Read: Keysight Introduces 1.6T Ethernet Test Platforms for Edge Routing and AI Data Centers
Highlighting the industry inflection point, Denise Dignam, President and CEO at Chemours, noted: “AI’s rapid growth is increasing the demands on data center infrastructure, and cooling is becoming a critical constraint. 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.” To accelerate broad market adoption, Chemours is actively collaborating across the infrastructure ecosystem with partners such as Oak Ridge National Laboratory, NTT DATA, 2CRSi, DataVolt, Munters, Accelsius, and ZutaCore.
