TPU regions and zones
Google Cloud uses regions, subdivided into zones, to define the geographic
location of physical compute resources. For example, the us-central1
region denotes a region near the geographic center of
the United States.
Google Cloud also offers AI zones which are specialized Google Cloud zones for AI/ML training and inference workloads. They provide significant ML accelerator (GPU/TPU) capacity. AI zones are geographically located typically within a 5ms radius from standard Google Cloud zones. For more information see About AI zones .
When you create a TPU VM, you specify the zone in which you want to create it. For more information about regional and zonal resources, see Compute Engine Global, regional, and zonal resources . You can create TPU VMs in the zones shown in the following table.
North America
-
us-central1-ai1 -
us-central1-b -
us-east1-d -
us-east5-a -
us-east5-b -
us-south1-ai1b
-
us-central1-a -
us-central1-b
-
us-central1-b -
us-central1-c
-
us-central1-a
-
us-central1-a
-
us-central1-a
-
us-central1-a
Europe
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
-
europe-west4-a
Asia Pacific
-
asia-east1-c
South America
TPU types with higher numbers of chips or cores are available only in limited quantities. TPU types with lower chip or core counts are more likely to be available.
What's next
- To see pricing for TPUs in each region, see Cloud TPU pricing .
- To learn how to work with TPUs, see Get started .
- To learn about TPU architecture, see TPU Architecture .
- To learn about the types of models that are well-suited to Cloud TPU, see When to use TPUs .
