Skip to content

Spin Up a Developer Pod

Step 1: Sign in to the Portal

Enter your email address or use one of the supported identity providers to sign in.

Sign in to the SHARON AI portal

Step 2: Open Developer Hub

After sign-in, you land on the main portal. From there, open the services launcher and choose Developer Hub or navigate via the Compute card.

Portal home after sign-in

Open Developer Hub from the services launcher

Step 3: Start a New Developer Pod

Inside Developer Hub, use the Developer Pods card to start the pod creation workflow. Select New Developer Pod to open the catalog.

Start a new Developer Pod from Developer Hub

Step 4: Select a Pod SKU

The Developer Pods Catalog lists available on-demand pod profiles with their GPU type and hourly cost. Select Select → on the desired SKU to continue.

Select a Pod SKU from the catalog

Step 5: Enter Pod Details

Complete the required fields on the New Developer Pod form:

  • Name: Friendly name for the pod.
  • Description: Optional description for the deployment.
  • Compute SKU: Preselected from the catalog step.
  • Workspace: Target workspace where the pod will be created.

The Container Configuration section appears below and requires selecting a Pod Image.

Enter the basic Developer Pod details

Step 6: Select a Container Image

Click the Pod Image dropdown to view all available container images. Options include Ubuntu base images and pre-built PyTorch images with specific CUDA versions.

Select a container image for the pod

Step 7: Review Details and Deploy

Confirm all settings look correct. The Monthly estimate panel on the right shows the expected cost before deployment. When ready, select Deploy.

Completed pod form ready to deploy

Step 8: Track Deployment Progress

After selecting Deploy, the platform navigates to the pod's deployment detail page. The status shows In Progress as compute resources are being allocated.

Note: You can navigate away from this page. The deployment continues in the background.

Pod deployment in progress — detail page

Pod deployment in progress — list view

Step 9: Deployment Success — Review Output

When provisioning completes, the status changes to Success. Open the pod's detail page and scroll to the Output section to collect the connection information:

  • Create Private Key File: Shell command to reconstruct the SSH private key locally.
  • SSH Command: Ready-to-run SSH command including the key, username, public IP, and port.

Copy the SSH command and use it in a terminal to connect to the pod.

Pod deployment success with output details

Step 10: Open Instance Actions and Select Delete

From the Developer Pods list, open the actions menu (...) for the instance and choose Delete.

Open pod actions and choose delete

Step 11: Confirm Pod Deletion

Type the exact instance name in the confirmation dialog and select Delete devpods.

Confirm pod deletion

Step 12: Deletion in Progress

After confirmation, the pod status changes to Destroying while deletion is in progress. The pod will disappear from the list once teardown completes.

Pod destroying status