> ## Documentation Index
> Fetch the complete documentation index at: https://docs.polystack.tech/llms.txt
> Use this file to discover all available pages before exploring further.

# Unified Console

> Manage CPU and GPU workloads, virtual machines, and Kubernetes clusters from one portal, with a single sign-on across the Polystack Dashboard and Rancher.

## Overview

The On-Prem AI Infrastructure Platform manages CPU and GPU workloads from **one console**.
The **Polystack Dashboard** handles virtual machines, flavors, and the Kubernetes cluster
lifecycle. **Rancher** handles management inside each cluster. Both sit behind one single
sign-on and are presented as one unified portal.

<Note>
  Every Kubernetes cluster created through [Polystack K8SaaS](/services/kubernetes/index) is
  imported into Rancher, so each new cluster appears in the portal without manual registration.
</Note>

***

## Console Components

| Component | Manages |
| - | - |
| **Polystack Dashboard** | Virtual machines, instance flavors, and the Kubernetes cluster lifecycle: create, scale, upgrade, delete |
| **Rancher** | In-cluster management of workloads, namespaces, and cluster resources across all imported clusters |
| **Single sign-on** | One login across both, backed by [Polystack Identity](/services/identity/index) and Keycloak |

***

## Architecture

```mermaid theme={null}
graph TD
    USER[User] --> SSO[Single Sign-On<br/>Polystack Identity + Keycloak]
    SSO --> PORTAL[Unified Portal]
    PORTAL --> DASH[Polystack Dashboard]
    PORTAL --> RANCHER[Rancher]
    DASH --> VMS[Virtual Machines and Flavors]
    DASH --> LC[Kubernetes Cluster Lifecycle]
    LC --> CLUSTERS[Kubernetes Clusters]
    CLUSTERS -->|imported| RANCHER
    RANCHER --> INC[In-Cluster Management]
```

***

## What Each Layer Handles

<Tabs>
  <Tab title="Infrastructure" icon="server">
    From the Polystack Dashboard:

    * Launch and manage CPU and GPU [virtual machines](/services/compute)
    * Select accelerator flavors for NVIDIA, Intel Gaudi, and AMD
    * Create, scale, upgrade, and delete [Kubernetes clusters](/services/kubernetes/index)
  </Tab>

  <Tab title="In-Cluster" icon="boxes-stacked">
    From Rancher:

    * Deploy and manage workloads across all clusters
    * Manage namespaces, services, and cluster resources
    * Work with every imported cluster from one view
  </Tab>

  <Tab title="Access" icon="user-shield">
    Through single sign-on:

    * One login for the Dashboard and Rancher
    * Identities managed through Polystack Identity and Keycloak
    * Consistent access across infrastructure and in-cluster operations
  </Tab>
</Tabs>

***

## Benefits

<CardGroup cols={2}>
  <Card title="One place for every workload" icon="layer-group" color="#bf9667">
    CPU and GPU virtual machines and Kubernetes clusters are managed from the same portal.
  </Card>

  <Card title="One login" icon="key" color="#bf9667">
    Single sign-on removes separate credentials for infrastructure and cluster management.
  </Card>

  <Card title="Automatic cluster onboarding" icon="plug" color="#bf9667">
    New Kubernetes clusters are imported into Rancher as they are created.
  </Card>

  <Card title="Clear separation of duties" icon="sitemap" color="#bf9667">
    Infrastructure lifecycle and in-cluster operations each have a dedicated, purpose-built
    interface.
  </Card>
</CardGroup>

***

## Next Steps

<CardGroup cols={2}>
  <Card title="On-Prem AI Infrastructure Platform" icon="brain" href="/services/ai-platform/index" color="#bf9667">
    Overview of all platform capabilities.
  </Card>

  <Card title="GPU-Capable VMs and Containers" icon="server" href="/services/ai-platform/gpu-vms-and-containers" color="#bf9667">
    The GPU workloads managed from the console.
  </Card>

  <Card title="Dashboard" icon="gauge" href="/services/dashboard/index" color="#bf9667">
    Using the Polystack Dashboard.
  </Card>

  <Card title="Identity and Access" icon="user-shield" href="/services/identity/index" color="#bf9667">
    Users, projects, and roles behind single sign-on.
  </Card>
</CardGroup>