GCP Compute Instance
tpu-v5p-32
The tpu-v5p-32 is a cloud compute GCP Compute instance with 0 vCPUs and 0 GB of memory, 32 TPU v5p GPUs, powered by Intel. Suitable for a broad range of cloud workloads including web serving, application hosting, and development environments. On-demand pricing starts at $134.4000/hr in us-central1. Spot pricing starts at $17.3032/hr in us-south1.
Price History
Pricing by Region
| Region | On-Demand ↑ | Spot | Reserved 1Y |
|---|---|---|---|
| us-central1 | $134.400 | $25.409 | $94.080 |
| us-east1 | $134.400 | $25.409 | $94.080 |
| us-east5 | $134.400 | $34.975 | $94.080 |
| europe-west4 | $136.080 | $37.733 | $95.424 |
| europe-west1 | $142.703 | $46.156 | $100.068 |
| us-south1 | $152.928 | $17.303 | $107.238 |
Spot Interruption Risk
Frequently Asked Questions
- What is the tpu-v5p-32?
- The tpu-v5p-32 is a cloud compute GCP Compute instance with 0 vCPUs and 0 GB of memory, 32 TPU v5p GPUs, powered by Intel. Suitable for a broad range of cloud workloads including web serving, application hosting, and development environments. On-demand pricing starts at $134.4000/hr in us-central1. Spot pricing starts at $17.3032/hr in us-south1.
- What is the cheapest region for tpu-v5p-32?
- us-central1 at $134.400/hr on-demand.
- How much does tpu-v5p-32 cost per month?
- Starting at $98112.00/mo on-demand in us-central1, or $12631.31/mo with spot pricing.