$ cat ABOUT.md

About CloudRig

Powerful computing shouldn't require expensive hardware. Here's the build log of why we made a raw GPU cloud for power users.

build_log
  1. 01[init]

    the problem

    A capable GPU workstation costs $2,000–$4,000 upfront, takes up space, generates heat, and depreciates the day you buy it. For developers, researchers, and creators, that cost is a barrier to doing the work they care about.

  2. 02[research]

    the gap

    Cloud GPU exists, but most of it is built for enterprises — buried in consoles, billed in opaque ways, or locked into proprietary notebooks. We wanted a raw box: a real NVIDIA GPU, root access, ready in under a minute, billed by the second.

  3. 03[build]

    what we shipped

    CloudRig is a thin, fast layer over a global GPU cloud. Pick a tier, deploy, ssh in. Full admin, persistent disk, connect from any device. When you're done, one command stops the meter and keeps your disk.

  4. 04[ship]

    the principle

    Pay only for what you use. Access it from anywhere. Never worry about upgrades again. No subscriptions, no contracts, no hardware to maintain.

infra
uptime_target
99.9%
boot_time
minutes
regions
12
billing
per-second
state_retention
24h
support_sla
<24h

// CloudRig runs on our GPU infrastructure partners' global cloud — data centers with high-performance NVIDIA GPUs, redundant networking, and fast storage. We target 99.9% uptime; affected users receive automatic credit for downtime. Machine state is preserved for 24 hours after stopping.

support
contact

$ need a hand with account, billing, or anything else?

→ support@cloudrig.space

// we respond within 24 hours

$ get started