KR6288

	Unprecedented AI Performance Powered by NVIDIA HGX H200 8-GPU 2x latest Intel® Xeon® or AMD EPYC™ processors 32 PFlops of industry-leading AI performance H200 Transformer Engine delivers supercharged training speed for GPT large language models
	Leading Architecture Design Lightning-fast CPU-to-GPU interconnect bandwidth Ultra-high scalable inter-node networking with up to 4.0 Tbps non-blocking bandwidth Optimized cluster-level architecture with 8:8:2 ratio of GPU to compute network to storage network
	Optimized Energy Efficiency Low air-cooled heat dissipation overhead and high power efficiency 54V, 12V separated power supply with N+N redundancy reducing power conversion loss Direct liquid cooling design with over 80% cold plate coverage keeps PUE ≤1.15
	Flexible Configurations for AI Scenarios Fully modular design and flexible configurations satisfy both on-premises and cloud deployment Easily harness large-scale model training, such as GPT-3, MT-NLG, stable diffusion and Alphafold Diversified SuperPod solutions accelerating the most cutting-edge innovation including AIGC, AI4Science and Metaverse