MQDB is NVIDIA’s blind mate quick coupling specification developed for the Vera Rubin architecture. As the blind mate version of the MQD series, it is designed for rack level liquid cooling systems in high density AI data centers and helps improve installation and maintenance efficiency in large scale deployments. Netonx MQDB complies with NVIDIA specifications and supports blind mating within the specified misalignment tolerance. It enables reliable connection and disconnection of node level manifolds without precise manual alignment. The flat face sealing structure and optimized flow path provide low fluid loss, high flow capability and low pressure drop.