Skip to content

Data center GPUs

NVIDIA B200 (HGX, SXM)

Memory

180 GB

Bandwidth

8,000 GB/s

FP16

2,250 TFLOPS

FP8

4,500 TFLOPS

Maker
NVIDIA
Model
B200 (HGX, SXM)
Type
Data center GPUs
Memory
180 GB
Memory type
HBM3e
Bandwidth
8,000 GB/s
FP16
2,250 TFLOPS
FP8
4,500 TFLOPS
FP4
9,000 TFLOPS
Power
1,000 W
Form factor
SXM
Interface
NVLink 5 1,800 GB/s; PCIe gen not stated per GPU
Released
2024

Notes on these figures

Per-GPU compute derived as HGX B200 8-GPU totals / 8 (NVIDIA publishes no per-GPU B200 table): FP16/BF16 36 PF and FP8 72 PF sparse (footnote: dense is half), FP4 72 PF dense. Memory 180GB HBM3e, up to 8TB/s, 1 kW per GPU configurable, NVLink 1800GB/s from https://docs.nvidia.com/enterprise-reference-architectures/hgx-ai-factory-h100-h200-b200/latest/components.html. Announced GTC March 2024.

Source: the maker's published specifications at nvidia.com. A missing figure means the maker does not publish it. Confirm current specifications with the maker or a dealer before you buy.

Comparable data center gpus

ModelMemoryMemory typeBandwidthFP16
AMD Instinct MI300X192 GBHBM35,300 GB/s1,300 TFLOPS
NVIDIA H200 NVL141 GBHBM3e4,800 GB/s835.5 TFLOPS
NVIDIA H200 SXM141 GBHBM3e4,800 GB/s989.5 TFLOPS
Intel Gaudi 3 (HL-325L OAM)128 GBHBM2E3,700 GB/s1,678 TFLOPS
Intel Gaudi 3 (HL-338 PCIe)128 GBHBM2E3,700 GB/s-
AMD Instinct MI325X256 GBHBM3E6,000 GB/s1,300 TFLOPS