NVIDIA DGX SuperPOD™ with DGX GB200 systems is purpose-built for training and inferencing trillion-parameter generative AI models. Each liquid-cooled rack features 36 NVIDIA GB200 Grace Blackwell Superchips–36 NVIDIA Grace CPUs and 72 Blackwell
GPUs–connected as one with NVIDIA NVLink. Multiple racks connect with NVIDIA Quantum InfiniBand to scale up to tens of thousands of GB200 Superchips.
•Highlyefficient, liquid-cooled, rack-scale design built with NVIDIA GB200 Grace
Hopper Superchips
•36NVIDIA Grace CPUs and 72NVIDIA Blackwell GPUs per rack, connected via fifth-generation NVLink
•Scaleto tens of thousands of GB200 Superchips with Quantum-2 InfiniBand
•Intelligent,full-stack resilience for constant uptime
•Integratedhardware and NVIDIA AI software
•Built,cabled, and factory tested before delivery and installation
•Optional 576 NVLinkconfiguration for memory-limited workloads
It is turnkey solutions providing end-to-end AI workload needed, from highperformance infrastructure to the MLOps for deployment.
