NVIDIA B200 Blackwell 
Specification NVIDIA B200
GPU Architecture NVIDIA Blackwell
GPU NVIDIA B200 Tensor Core GPU
GPU Memory 180GB HBM3e
Memory Bandwidth Up to 8 TB/s
Form Factor SXM6
GPU Interconnect 5th Generation NVLink
NVLink Bandwidth Up to 1.8 TB/s per GPU
TDP Up to 1,000W
Tensor Cores 5th Generation
RT Cores Blackwell RT Core architecture
MIG Up to 7 GPU Instances
PCIe Server platform dependent
Display Outputs None
ECC Yes
Memory Type HBM3e
Target Market Data Center / AI / HPC
Typical Platform HGX B200 / DGX B200 / GB200
Typical HGX Configuration 8 × B200 GPUs