FusionPoD & Liquid Cooling

FusionPoD & Liquid Cooling

FusionPoD for AI delivers high-density AI infrastructure through integrated-cabinet design, 100% native liquid cooling and bus blind-mating technology.

FusionPoD for AI

Rack-scale AI, native liquid cooling

Verified highlights

• Up to 64 GPUs
• Cabinet pPUE as low as 1.06
• 105 kW for centralized, efficient power supply
• 2-bus blind mating
• Large-model training, data analysis and HPC; telecom, Internet, finance and government/enterprise uses

FusionPoD liquid-cooled server

FusionPoD is a high-density, energy-efficient rack-scale platform with integrated delivery for cloud, HPC and liquid-cooled data-center scenarios.

Highlights

• pPUE as low as 1.06 for energy efficiency
• Suited to cloud, virtualization, big data and HPC
• Integrated delivery for IDC, telecom and Internet data centers

Product specifications may change by configuration or region. For current details visit the official page: xfusion.com/en/product/rack-scale-servers/fusionpod-for-ai