Skip to content

H200 SXM vs H100 SXM

Side-by-side comparison of the NVIDIA H200 SXM and the NVIDIA H100 SXM for AI inference workloads.

Specifications

SpecH200 SXMH100 SXM
Generationhopperhopper
Memory TypeHBM3eHBM3
VRAM141 GB80 GB
Memory Bandwidth4,800 GB/s3,350 GB/s
BF16 TFLOPS990990
FP16 TFLOPS990990
FP8 TFLOPS1,9791,979
INT8 TOPS1,9791,979
TDP700 W700 W
Interconnectnvlinknvlink
NVLink Bandwidth900 GB/s900 GB/s
Max GPUs per Node88
PCIe GenGen 5Gen 5
CUDA Compute Capability99

Pricing

H200 SXM

ProviderOn-DemandReservedSpot
lambda$3.49/hr$2.69/hr-
coreweave$4.25/hr$3.19/hr-
runpod$4.69/hr--
tensordock$3.80/hr-$2.90/hr

H100 SXM

ProviderOn-DemandReservedSpot
runpod$4.18/hr-$3.29/hr
lambda$2.49/hr$1.89/hr-
coreweave$3.79/hr$2.57/hr-
aws$5.12/hr$3.59/hr-
gcp$4.85/hr$3.40/hr-
azure$4.98/hr$3.49/hr-
vast ai$3.40/hr-$2.50/hr
tensordock$3.29/hr-$2.49/hr
fluidstack$2.85/hr-$2.10/hr

Cheapest available rate: H200 SXM at $2.69/hr vs H100 SXM at $1.89/hrH100 SXM is +42% cheaper

Efficiency Metrics

TFLOPS / Watt

1.4

H200 SXM

vs

1.4

H100 SXM

BF16

VRAM / Dollar

52.4

H200 SXM

vs

42.3

H100 SXM

GB/$/hr

Bandwidth / Watt

6.9

H200 SXM

vs

4.8

H100 SXM

GB/s/W

Models (FP16, 1 GPU)

193.0

H200 SXM

vs

182.0

H100 SXM

Model Compatibility (FP16, Single GPU)

Only on H200 SXM (11)

  • Jamba 1.5 Mini
  • Amazon Nova Pro
  • Falcon 40B
  • Mixtral 8x7B
  • Llama 3.1 Nemotron 51B
  • VILA 1.5 40B
  • Gemini 2.0 Flash
  • Gemini 1.5 Flash
  • Jamba Instruct
  • Phi 3.5 MoE
  • Mixtral 8x7B Instruct

Both (182)

  • Yi 1.5 34B
  • Yi 1.5 9B
  • Yi Coder 9B
  • GTE Qwen2 7B
  • Marco O1
  • Qwen 1.5 MoE A2.7B
  • Qwen 2 Audio 7B
  • Qwen 2.5 14B
  • Qwen 2.5 32B
  • Qwen 2.5 3B
  • Qwen 2.5 Coder 32B
  • OLMo 2 13B
  • OLMo 2 7B
  • Amazon Nova Lite
  • OpenELM 3B
  • BGE Large EN v1.5
  • BGE M3
  • Baichuan 2 13B
  • OctoCoder 15B
  • StarCoder2 15B
  • +162 more

Only on H100 SXM (0)

None

Summary

The H200 SXM (hopper generation) offers 141GB of HBM3e with 990 BF16 TFLOPS and 4,800 GB/s memory bandwidth at 700W TDP.

The H100 SXM (hopper generation) offers 80GB of HBM3 with 990 BF16 TFLOPS and 3,350 GB/s memory bandwidth at 700W TDP.

The H200 SXM has +76% more VRAM, allowing it to run larger models without multi-GPU setups.

From a cost perspective, the H100 SXM is more affordable at $1.89/hr vs $2.69/hr for the H200 SXM.

More GPU Comparisons