The Trillion-Parameter Instrument of AI

Open up enormous potential in the age of AI with a new class of AI supercomputer that fully connects 256 NVIDIA Grace Hopper™ Superchips into a singular GPU. NVIDIA DGX™ GH200 is designed to handle terabyte-class models for massive recommender systems, Generative AI, and graph analytics, offering 144 terabytes (TB) of shared memory with linear scalability for giant AI models.

→ Giới thiệu chi tiết NVIDIA DGX GH200

→ Datasheet DGX GH200

SPECIFICATIONS

CPU and GPU

256x NVIDIA Grace Hopper Superchips

CPU Cores 

18,432 Arm® Neoverse V2 Cores with SVE2 4X 128b

GPU Memory 

144TB

Performance

1 exaFLOPS

Networking

• 256x OSFP single-port NVIDIA ConnectX®-7 VPI with 400Gb/s InfiniBand
• 256x dual-port NVIDIA BlueField®-3 VPI with 200Gb/s InfiniBand and Ethernet
• 24x NVIDIA Quantum-2 QM9700 InfiniBand Switches
• 20x NVIDIA Spectrum™ SN2201 Ethernet Switches
• 22x NVIDIA Spectrum SN3700 Ethernet Switches

NVIDIA NVLink Switch System 

• 96x L1 NVIDIA NVLink Switches
• 36x L2 NVIDIA NVLink Switches

Software

• NVIDIA AI Enterprise (optimized AI software)
• NVIDIA Base Command (orchestration, scheduling, and cluster management)
• DGX OS / Ubuntu / Red Hat Enterprise Linux / Rocky (operating system)

Management Network 

Host baseboard management controller (BMC) with RJ45

Support

Comes with three-year business-standard hardware and software support