Deployment

We deliver the AI Platform as a virtual appliance: a pre-built virtual machine image you run on your own infrastructure. Installation takes four stages, from deploying the appliance to connecting it to the cloud Console, where you register, diagnose, and upgrade your appliances.

flowchart LR
    A["Deploy the<br/>Virtual Appliance"] --> B["Configure the<br/>Virtual Appliance"]
    B --> C["Register the<br/>Appliance"]
    C --> D["Upgrade the<br/>Appliance<br/>(if available)"]

    classDef required fill:#e8f0fe,stroke:#1a73e8,stroke-width:2px,color:#174ea6
    classDef optional fill:#f1f3f4,stroke:#9aa0a6,stroke-width:2px,color:#3c4043,stroke-dasharray:5 3

    class A,B,C required
    class D optional

Deployment Targets

We ship a separate image for each target platform. Choose the one that matches your infrastructure:

Machine Specifications

The minimum machine specifications are comparable to an AWS c6i.xlarge EC2 instance:

Resource Minimum
Memory 16 GB
CPU 4 vCPUs
Disk 64 GB
Note

Sizing Consideration

On most virtualization platforms, you can resize memory and CPU after deployment. Disk resizing is more involved and varies by platform, so allocate enough storage up front.

Step 1: Deploy the Virtual Appliance

Deploy the virtual appliance from a virtual machine image onto your infrastructure, creating the virtual machine that runs the platform.

Step 2: Configure the Virtual Appliance

On first boot, configure the appliance by setting the time zone, network, hostname, and DNS so it has a static address and can join your network and be reachable.

Step 3: Register the Appliance

Register the appliance by generating a license and connecting it to the cloud Console. Registration is required before you can use the appliance.

Step 4: Upgrade the Appliance

When an upgrade is available, upgrade the appliance to bring its component services to the latest version. This stage is optional and applies only when a newer version exists. You manage and run upgrades from the cloud Console.