NVIDIA HGX H200

baseboard
8× H200
NVSwitch + NVLink
2024
FP64
0.3
PFLOPS
FP32
0.5
PFLOPS
Power
8.0
kW Total
Memory
1128
GB Total

HGX platform for cloud providers with H200 GPUs

FP64
0.27
PFLOPS
68.0 TFLOPS/kW
FP32
0.54
PFLOPS
134.0 TFLOPS/kW
TF32
3.96
PFLOPS
989.0 TFLOPS/kW
FP16
7.92
PFLOPS
1979.0 TFLOPS/kW

System Details

GPU Configuration

GPU Count: 8 GPUs
Architecture: Hopper
Interconnect: NVSwitch + NVLink

System Specifications

Form Factor: baseboard
Total Power: 8.0 kW
Total Memory: 1128 GB
Memory Bandwidth: 38400 GB/s

Precision Performance Breakdown

PrecisionSystem PerformancePer GPUEfficiency
FP64 0.272 PFLOPS 34.0 TFLOPS 68.0 TFLOPS/kW
FP32 0.536 PFLOPS 67.0 TFLOPS 134.0 TFLOPS/kW
TF32 3.956 PFLOPS 494.5 TFLOPS 989.0 TFLOPS/kW
FP16 7.916 PFLOPS 989.5 TFLOPS 1979.0 TFLOPS/kW
BF16 7.916 PFLOPS 989.5 TFLOPS 1979.0 TFLOPS/kW
FP8 15.832 PFLOPS 1979.0 TFLOPS 3958.0 TFLOPS/kW
INT8 15.832 PFLOPS 1979.0 TFLOPS 3958.0 TFLOPS/kW

Powered by NVIDIA H200

This system utilizes 8 × NVIDIA H200 SXM 141GB GPUs, each delivering exceptional performance for AI and HPC workloads.

Per GPU TDP

700W

Per GPU Memory

141 GB

Process Node

4nm

Architecture

Hopper

Documentation & Resources

workstation-datasheet-dgx-spark-gtc25-spring-nvidia-us-3716899-web.pdf

NVIDIA • 2025-10-16

View Document ↗

NVIDIA DGX Documentation

System guides, deployment resources

Browse ↗

Typical Use Cases

Large Language Model Training
Distributed Deep Learning
Multi-GPU Inference
HPC Simulations
Scientific Computing

The NVIDIA HGX H200 with 8× H200 GPUs is designed for enterprise-scale workloads requiring 0.3 PFLOPS of compute power with NVSwitch + NVLink interconnect for efficient multi-GPU communication.