How to Run a GPU-Accelerated Virtual Desktop Environment on RunPod

BY ZHEN LU - JUNE 28, 2022
There are a few use cases for running a GPU accelerated virtual desktop on RunPod. One common use case is for those who need to use demanding applications that require a lot of processing power, such as 3D rendering with blender. In this blog post, we'll go over how to easily deploy a pod with everything you need to get rendering.

In this post, we show you how to run Ubuntu Remote desktop with access to GPU running on RunPod.

First, make sure you have some funds in your account.

Then, you can use this link to deploy the latest desktop template: Click Here
Pick the GPU you want to use and start your pod using on-demand.

Default username: kasm_user

Default password: password

While you can use the template with the default password, you may want to secure it a bit further by changing it to something better.

To do this click Edit Pod

Then head over to Environment Variables section and replace password with password of your choice.

After that apply setting and pod should start. Give it some time to fully load till the Connect button becomes active.

Then you will be able to connect to your remote desktop using the "Connect via HTTP" button.

If you seeing error 502 or 404 that's probably because the Pod is not fully started. Give it a minute and try it again. A good indication that the pod is ready is the CPU utilization going to 0%.

The browser will ask you to provide username and password to login
The username is always: kasm_user
The password is based what you set in Environment Variables section

After entering password you should see your desktop screen.

Important notes:
This template is set to save /home as workspace so any preinstalled apps will be gone on Pod reset, but will be retained on pod restart.
At the middle of the left side there is a button to open a side bar where you can find additional access to tools like clipboard or settings to control streaming quality.

Zhen is one of the co-founders of RunPod. He holds a PhD in Computational Chemistry and has lots of experience teaching, coding, and talking to people.

Want to join our growing community?
Chat with us on our Discord Server.
Prefer email? Contact us at