PyyanThe AI Index

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

TPU v7 Ironwood vs B300 Blackwell Ultra vs H100

3 of 5

GPU & AI Chips · verified 13 Aug 2026

×TPU v7 IronwoodGooglecurrent
×B300 Blackwell UltraNVIDIAcurrent
×H100NVIDIAcurrent
2 slots left
SpecificationTPU v7 IronwoodB300 Blackwell UltraH100
SummaryInference-oriented, Google Cloud only.15 PFLOPS dense FP4.The workhorse of the previous generation.
Memory192GB HBM3e288GB HBM3e80GB HBM3
Bandwidth7.37 TB/s8 TB/s—
Price—~$40–50K (verify)$25–40K
VendorGoogleNVIDIANVIDIA
CategoryGPU & AI ChipsGPU & AI ChipsGPU & AI Chips
OfficialGoogle ↗NVIDIA ↗NVIDIA ↗

Highlighted rows are where these differ.

TPU v7 Ironwood

  • Requires JAX
  • Not purchasable or on any marketplace

Best for GCP inference.

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 →

H100

The workhorse of the previous generation.

Best for value.

Full spec sheet →