ComputeAI Servers
AI Servers
Dedicated NVIDIA GPU servers with drivers pre-installed, for rendering, VFX, training and inference, billed by the hour while they run.
An AI server is a dedicated GPU server: one or more NVIDIA GPUs, with the drivers pre-installed and full root (or Administrator) access. It works like an instance, with GPU plans and GPU metrics.
| AI Servers | Instances | |
|---|---|---|
| Hardware | Dedicated NVIDIA GPUs plus CPU, RAM and SSD | Virtual CPU, RAM and SSD |
| Good for | Rendering, VFX, training, inference | Apps, websites, APIs |
| Connect | SSH, or Remote Desktop on Windows | SSH, or Remote Desktop on Windows |
| Where | Compute › AI Servers | Compute › Instances |
Choose a GPU package
Compare GPU packages by GPUs, VRAM, CPU, RAM, disk and network, and see their prices per hour, day, week or month before you deploy one.
Create an AI server
Launch a dedicated GPU server in minutes. Pick a GPU plan, an operating system and a location, choose how you log in and how you're billed.
Connect to an AI server
Log in to a GPU server over SSH on Linux or Remote Desktop on Windows, use the browser console, and move files over SFTP.
Manage an AI server
Power, networking, resizing, backups, billing and deletion work the same on AI servers as on instances, with GPU metrics and GPU-only resizing.
Rename, move or delete a server
Rename a cloud server, move it to another project so its costs show there, reset the root password, or delete the server and stop billing.
Choose a GPU package
Compare GPU packages by GPUs, VRAM, CPU, RAM, disk and network, and see their prices per hour, day, week or month before you deploy one.