PyyanThe AI Index

Pyyan / Compare / H100 vs B300 Blackwell Ultra vs TPU v7 Ironwood

H100 vs B300 Blackwell Ultra vs TPU v7 Ironwood

3 of 5

GPU & AI Chips · verified 13 Aug 2026

×H100NVIDIAcurrent
×B300 Blackwell UltraNVIDIAcurrent
×TPU v7 IronwoodGooglecurrent
2 slots left
SpecificationH100B300 Blackwell UltraTPU v7 Ironwood
SummaryThe workhorse of the previous generation.15 PFLOPS dense FP4.Inference-oriented, Google Cloud only.
Memory80GB HBM3288GB HBM3e192GB HBM3e
Bandwidth—8 TB/s7.37 TB/s
Price$25–40K~$40–50K (verify)—
VendorNVIDIANVIDIAGoogle
CategoryGPU & AI ChipsGPU & AI ChipsGPU & AI Chips
OfficialNVIDIA ↗NVIDIA ↗Google ↗

Highlighted rows are where these differ.

H100

The workhorse of the previous generation.

Best for value.

Full spec sheet →

B300 Blackwell Ultra

  • 8 TB/s memory bandwidth
  • Per-GPU pricing widely misquoted — one source cites rack pricing

Best for frontier training.

Full spec sheet →

TPU v7 Ironwood

  • Requires JAX
  • Not purchasable or on any marketplace

Best for GCP inference.

Full spec sheet →