Machines are high-performing computing for scaling AI applications.
Paperspace offers a wide range of compute machines that you can manage from the console or API.
The key building blocks are machines that can be scaled up or down and can be configured with various storage and networking options. Machines are Linux or Windows-based virtual machines that run on top of virtualized hardware. Each machine you create is a new server or desktop you can use, either standalone or as part of a larger, cloud-based infrastructure.
Machines achieve near bare-metal performance. Machines are multi-tenant which means they run on shared hardware but are completely isolated at the hypervisor level. Paperspace provides each machine with a consistent and predictable amount of CPU, RAM, and GPU capacity, regardless of its underlying hardware. Machines use persistent storage, ensuring a familiar experience, commensurate with local computing.
Core includes:
Multiple OS options as well as custom templates
A large set of GPU machines
Global availability with three datacenter regions around the globe and 99.99% uptime
User management with Teams
Free unlimited bandwidth
An API for automating tasks
Paperspace allows you to set an autoshutdown time limit on all machine types. This is to prevent unnecessary account charges for long running Gradient processes. The autoshutdown timer can be set between 1 hour and 1 week. You may remove the limit but this should be closely monitored because it can lead to large utilization charges.
When using free-tier machines, the maximum autoshutdown timeout is 6 hours. This ensures we maintain fairness and capacity for the entire Paperspace community.