Pyyan / Compare / H100 vs B200 vs MI400 vs TPU v7 Ironwood

H100 vs B200 vs MI400 vs TPU v7 Ironwood

4 of 5

GPU & AI Chips · verified 13 Aug 2026

×H100NVIDIAcurrent
×B200NVIDIAcurrent
×MI400AMDcurrent
×TPU v7 IronwoodGooglecurrent
1 slot left
SpecificationH100B200MI400TPU v7 Ironwood
SummaryThe workhorse of the previous generation.The current volume Blackwell part.Double NVIDIA's memory capacity.Inference-oriented, Google Cloud only.
Memory80GB HBM3192GB HBM3e432GB HBM4192GB HBM3e
Bandwidth19.6 TB/s7.37 TB/s
Price$25–40K$30–50K
VendorNVIDIANVIDIAAMDGoogle
CategoryGPU & AI ChipsGPU & AI ChipsGPU & AI ChipsGPU & AI Chips
OfficialNVIDIANVIDIAAMDGoogle

Highlighted rows are where these differ.

H100

The workhorse of the previous generation.

Best for value.

Full spec sheet →

B200

  • Sold out through mid-2026
  • ~3.6M unit backlog

Best for general training.

Full spec sheet →

MI400

  • 19.6 TB/s bandwidth
  • 40 PFLOPS FP4
  • $7.2B in commitments

Best for large-model inference.

Full spec sheet →

TPU v7 Ironwood

  • Requires JAX
  • Not purchasable or on any marketplace

Best for GCP inference.

Full spec sheet →