AI Zones

AI zones are specialized zones used for Artificial Intelligence and Machine Learning (AI and ML) training and inference workloads. They provide significant ML accelerator (GPU and TPU) capacity.

Within a region, AI zones are geographically located away from standard (non-AI) zones. The following figure shows an example of an AI zone (us-central1-ai1a) located further away relative to the standard zones in the us-central1 region.

Note: AI zones meet the security and data residency requirements of their region, similar to the standard zones. For more information, see Geographic management of data.

Parent zone

Each AI zone is associated with a standard zone in the region, referred to as its parent zone. A parent zone is a standard zone with the same suffix as the AI zone. For example, in the diagram, us-central1-a is the parent zone of us-central1-ai1a. They share software update schedules and sometimes infrastructure. This means that any software or infrastructure issues affecting a parent zone could also affect the AI zone. When designing your high availability solutions, review the High availability (HA) considerations to account for the dependency on the parent zone.

When to use AI zones

AI zones are optimized for AI and ML workloads. Use the following guidance to determine which of your workloads are best suited for AI zones and which are better served by standard zones.

Recommended for:

• Large-scale training: Ideal for large-scale training workloads—such as Large Language Model (LLM) and foundational model training—because of the availability of a large number of accelerators.

• Small-scale training, fine-tuning, bulk inference, and retraining: AI zones perform well for workloads that require substantial accelerator capacity.

• Real-time ML inference: AI zones support real-time inference workloads. Performance depends on the application design and model latency requirements, especially if the workload requires round-trip requests to the parent region.

Not recommended for:

• Non-ML workloads: Since AI zones do not offer all Google Cloud services locally, we recommend running your non-ML workloads in the standard zones.

Access services from an AI zone

You can access all Google Cloud products in a Google Cloud region from its AI zone. However, accessing services in a Google Cloud region from an AI zone can add network latency, as the AI zone is physically separate from the locations of the region's standard zones.

Specific products support creating or accessing zonal resources locally in an AI zone. For more information about these services, see the following table:

Product	Description
Google Kubernetes Engine (GKE)	Setup for using AI zones in GKE clusters, including configuration using ComputeClasses, node auto-provisioning, and GKE Standard node pools. Using AI zones in GKE
Cloud Storage	Configuration of object storage for workloads in AI zones, including zonal storage to maximize performance during active jobs and persistent storage for datasets and model checkpoints. Use AI zones with Cloud Storage
Compute Engine	Methods to identify available AI zones using the console, Google Cloud CLI, and REST API, including how to filter by naming convention, accelerator type, or machine Find available AI zones

Locations

AI zones are available in the following locations:

AI zone	AI zone location	Google Cloud region	Google Cloud region location	Parent zone
us-south1-ai1b	Austin, Texas, North America	us-south1	Dallas, Texas, North America	us-south1-b
us-central1-ai1a	Lincoln, Nebraska, North America	us-central1	Council Bluffs, Iowa, North America	us-central1-a

Using AI zones

AI zones are accessible through the Google Cloud console, Google Cloud CLI, or REST. However, when using the Google Cloud console to create your VMs, you must manually select an AI zone. It isn't selected for you, as it is with standard zones. To use AI zones with the following features, you must explicitly select an AI zone while you are setting up these resources.

• Certain Compute Engine and GKE features: AI zones are not automatically selected in certain Compute Engine and GKE regional features (for example, Regional Managed Instance Groups, Regional GKE clusters). For more details about GKE, refer to the GKE documentation.

• Non-accelerator workload restrictions: When you run CPU-only VMs in AI zones, be aware of Compute Engine-enforced restrictions. These might include requirements for GPU:CPU ratios and reservations.

• Vertex AI: GKE based Vertex AI regional products must configure GKE to include AI zones in regional clusters. You don't need to opt in to Vertex AI. Vertex AI manages this configuration.

• Google Cloud Service Metadata Locations API: You must enable the --extraLocationTypes flag when using the locations.list API to ensure AI zones appear only to those who intend to use them.

Using AI zones in GKE

By default, GKE doesn't deploy your workloads in AI zones. To use an AI zone, you configure one of the following options:

• ComputeClasses: Set your highest priority to request on-demand TPUs in an AI zone. ComputeClasses help you define a prioritized list of hardware configurations for your workloads. For an example, see About ComputeClasses.

• Node auto-provisioning: Use a nodeSelector or nodeAffinity in your pod specification to instruct node auto-provisioning to create a node pool in the AI zone. If your workload doesn't explicitly target an AI zone, node auto-provisioning considers only standard zones when creating new node pools. This configuration ensures that workloads that don't run AI/ML models remain in standard zones unless you explicitly configure otherwise. For an example of a manifest that uses a nodeSelector, see Set the default zones for auto-created nodes.

• GKE Standard: If you directly manage your node pools, use an AI zone in the --node-locations flag when you create a node pool. For an example, see Deploy TPU workloads in GKE Standard.

Limitations

The following are not available in AI zones:

• Regional persistent disks
• Hyperdisk Balanced High Availability

Design considerations with AI zones

Consider the following when designing your applications to use AI zones.

High availability (HA) considerations

AI zones share software rollouts and infrastructure with their parent zones. To ensure high availability for your workloads, avoid these deployment patterns when you select zones, whether automatically or manually:

• Avoid deploying HA workloads across an AI zone and its parent zone.

• Avoid deploying HA workloads across two AI zones that share the same parent zone.

Storage best practices

We recommend a tiered storage architecture to balance cost, durability, and performance:

1. Cold storage layer: Use regional Cloud Storage buckets in standard zones for persistent, highly-durable storage of your training datasets and model checkpoints.

2. Performance layer: Use specialized zonal storage services to act as a high-speed cache or temporary scratch space. This approach eliminates inter-zonal latency and maximizes goodput during active jobs.

   To help ensure that GPUs and TPUs remain fully saturated, maximizing goodput, provision your performance layer in the same AI zone as your compute resources.

The following storage solutions are recommended for optimizing AI and ML system performance with AI zones:

Storage service	Description	Use cases
Anywhere Cache feature of Cloud Storage	A fully managed, SSD-backed zonal read cache that brings frequently read data from a bucket into the AI zone.	Recommended for: Read-heavy workloads Low-latency model training and serving Not recommended for: Applications that require full POSIX compliance

What's next

• View a list of AI zones
• View TPU types and locations available in AI zones
• Set up an organization policy to restrict VM creation in AI zones