Description of image

Machines How-Tos

Machines are high-performing computing for scaling AI applications.

Getting Started

Create a machine in the Paperspace console.
Connect to your compute machines with SSH.
Connect to your Windows machine with the desktop app.


Add more CPU, RAM, GPU, and storage to your machine.
Create a custom template of a Paperspace machine.
Create manual and automatic backups of your compute machines.
Prevent auto-shutdown on your Linux Ubuntu machine while connecting via SSH.
Disable NVLink to ensure CUDA runs on your machines.
Configure NVLink for non-ML-in-a-Box or terminal/ssh-only (headless) machines.

Manage Data

Load data from various sources onto your machine.
Shared storage accessible from machines within a private network.
Load files from your virtual machine (VM) using FileZilla Client and FileZilla Server.
Mount shared drives on Windows and Linux machines.
Load files from your virtual machine (VM) usingWindows’ RDP.


Attaching a public IP makes your compute machines addressable over the internet.
Use a private network (a dedicated subnet) to secure communication between machines and shared drives.
Configure a VPN tunnel between your Paperspace machine and any other computer.
Enable two-way communication between your Paperspace machines and devices in your office or third-party site.
Map hostnames to IP addressing by editing the Windows hosts file.
Persist a hostname through reboots and stopping or starting your instance with a static hostname.
Test latency to Paperspace regions.


Manage access to Paperspace machines.
Information about credentials and access to machines.