io.net primarily supports Linux workloads (Ubuntu, Debian, CentOS) for GPU computing. Windows workloads can run via Docker containers with WSL2 (Windows Subsystem for Linux) support on select nodes, or through virtual machines for specific enterprise use cases—contact support for Windows availability.

Linux (Primary Support)

• Ubuntu 20.04/22.04 (recommended)
• Debian, CentOS, Rocky Linux
• All CUDA versions supported

Windows Support

Limited availability:
- Windows Server 2019/2022 on select nodes
- RDP access for remote desktop
- Use case: Windows-specific ML frameworks, DirectML

Request Windows node:

io deploy --os windows-server-2022 --gpu RTX4090
# Availability varies, may require enterprise account

Alternative: Docker + WSL2

Run Windows containers on Linux hosts:

FROM mcr.microsoft.com/windows/servercore:ltsc2022
# Windows-based application

Most users run Linux workloads (99%+ of GPU ML/AI workloads are Linux-based).

Linux-first GPU cloud — io.net optimized for ML/AI (Windows via Docker/enterprise).