NVIDIA Google TPU

TPU v5 Lite Pod (8 cores)

Eight-core v5 Lite pods provide substantial throughput for inference-heavy production deployments.

Launch year
2024
Memory
256 GB HBM
Memory bandwidth
>900 GB/s
Peak FP16 / FP32
480 TFLOPS · 0 TFLOPS

Market snapshot

$4.08 /hr

Range $4.08 – $12.48

Catalog coverage

32 live offerings

Across 1 providers · 18 regions

  • Scales via TPU pod fabric
  • Low-latency inference
  • Cloud-native management
TPU v5 Lite Pod (8 cores) render

Last refreshed Oct 20, 2025, 2:01 AM

Performance snapshot

Normalized versus NVIDIA A100 (=1.0). Values use public reference benchmarks for training and inference workloads.

  • TPU Throughput×0.85
  • Energy Efficiency×0.85
  • Latency×0.60

Provider availability

Price bands per provider pulled from the live catalog.

  • Google Cloud logoGoogle Cloud32 offers
    $4.08 – $12.48 /hr

Popular regions

  • us-west4-b2 offers
  • us-west4-a2 offers
  • us-east5-a2 offers
  • us-east5-c2 offers
  • us-east5-b2 offers
  • us-west1-c2 offers
  • us-central1-a2 offers
  • europe-west1-c2 offers

Weighted average price $8.30 /hr · median $9.60