The FusionPoD for AI rack-scale liquid-cooled server adopts the industry-leading integrated cabinet design and 100% native liquid cooling and bus blind mating technologies, delivering a peak computing with 64 GPUs. It supports automatic inspection with robots and unmanned O&M, and features high computing, high performance, high energy efficiency, high reliability, high bandwidth, and simplified O&M.
Application scenarios
- Large model training, data analysis, and high performance computing, Telecom, Internet, finance, and government and enterprise
