Pyyan / Compare / B300 Blackwell Ultra vs MI400 vs TPU v7 Ironwood

B300 Blackwell Ultra vs MI400 vs TPU v7 Ironwood

3 of 5

GPU & AI Chips · verified 13 Aug 2026

×B300 Blackwell UltraNVIDIAcurrent
×MI400AMDcurrent
×TPU v7 IronwoodGooglecurrent
2 slots left
SpecificationB300 Blackwell UltraMI400TPU v7 Ironwood
Summary15 PFLOPS dense FP4.Double NVIDIA's memory capacity.Inference-oriented, Google Cloud only.
Memory288GB HBM3e432GB HBM4192GB HBM3e
Bandwidth8 TB/s19.6 TB/s7.37 TB/s
Price~$40–50K (verify)
VendorNVIDIAAMDGoogle
CategoryGPU & AI ChipsGPU & AI ChipsGPU & AI Chips
OfficialNVIDIAAMDGoogle

Highlighted rows are where these differ.

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 →

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 →