Fabric8Labs announced its collaboration with Wiwynn to develop and showcase next-generation ECAM-enabled cold plate solutions. The advanced technology will be showcased at OCP Global Summit 2025, taking place October 13-16 at the San Jose Convention Center. At Wiwynn's booth, attendees will see a range of cutting-edge cold plate solutions manufactured with Fabric8Labs' proprietary ECAM process, including: Double-sided cold plates that are designed and parametrically optimized to extract heat from both sides of Future AI modules.

Two-phase cold plates with advanced wicking structures developed by Fabric8Labs in collaboration with Wiwynn. A next-generation cold plate presentation with ECAM microchannels for ultra-high heat fluxes beyond 350W/cm2 at the Future Technologies Symposium. This joint effort highlights the transformative potential of Electrochemical Additive Manufacturing (ECAM), a process that enables the creation of complex, high-performance data center cooling hardware.

Fabric8Labs' ECAM process unlocks new capabilities in thermal management design by manufacturing highly customized, complex geometries that are optimized to a chip's unique power map. When integrated into the AI server's cold plates, the technology delivers superior thermal performance that is essential for cooling the high-wattage GPUs and ASICs in modern AI data centers. The demonstrated hardware solutions highlight a significant leap in cooling technology, driven by ECAM-enabled cold plates.

These cold plates provide: Superior Thermal Performance: Enabling optimized fin structures that precisely direct coolant towards the chip's hot spots, these cold plates maximize cooling performance where it's needed most, potentially resulting in a 5-8degC reduction in maximum chip temperature for high TDP devices. Enhanced Temperature Uniformity: ECAM cold plates with a customized microchannel design precisely directing coolant flow, significantly reducing temperature fluctuations, resulting in a 200% improvement in temperature variance across the chip. This uniformity is crucial for maximizing chip performance and longevity.

Reduced Total Cost of Ownership (TCO): By integrating best-in-class thermal performance and improved reliability, Wiwynn's IT infrastructure enables lower operational costs for AI data centers through reduced spending on refrigeration and pumping power to move coolant. Together, Fabric8Labs and Wiwynn are redefining thermal management for AI infrastructure--delivering scalable, efficient, and sustainable solutions that push the boundaries of what's possible in data center cooling.