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 $21.3504/hr in us-south1.
Price History
Pricing by Region
| Region | On-Demand ↑ | Spot | Reserved 1Y |
|---|---|---|---|
| us-central1 | $134.400 | $28.430 | $94.080 |
| us-east1 | $134.400 | $28.430 | $94.080 |
| us-east5 | $134.400 | $32.108 | $94.080 |
| europe-west4 | $136.080 | $32.825 | $95.424 |
| europe-west1 | $142.703 | $53.834 | $100.068 |
| us-south1 | $152.928 | $21.350 | $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 $21.3504/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 $15585.77/mo with spot pricing.