# Ardus

> Ardus finds existing, powered data center capacity from 1 MW to 10 MW. Customers map their compute requirements; Ardus agents search sites, verify power, cooling, fiber, and access, rank them by months to first MW, and set a go-live date. Customers bring only their GPUs. Ardus is operated by Standard Inference, Inc. Contact: hello@ardus.ai.

Who it is for: neoclouds with GPUs already on order, AI labs, enterprise teams bringing their own GPUs, and inference companies expanding region by region.

To start a project, send requirements (MW, rack density, cooling, redundancy, region or latency target, go-live date, GPU arrival, term and budget) via https://ardus.ai/contact (the form then offers a 30-minute call, also bookable directly at https://cal.com/ardus/30min) or email hello@ardus.ai. After a first call, customers ready to move forward sign an NDA, complete a credit application, and sign an LOI (non-binding or binding, depending on the deal). Above 10 MW, Ardus scopes it with the customer's team.

Every page below is also available as Markdown. The full text of all pages is at https://ardus.ai/llms-full.txt.

## Product

- [Overview](https://ardus.ai/index.md): What Ardus does, how it works, and what it covers
- [Requirements](https://ardus.ai/features/modeling.md): Requirements in; site, go-live date, critical path, and terms out
- [Sites](https://ardus.ai/features/sites.md): Existing sites with power and cooling available now, verified and ranked by months to first MW
- [Move in](https://ardus.ai/features/move-in.md): Terms, access, fiber, fit-out, and go-live tracking at an existing site

## Blog

- [Introducing Ardus](https://ardus.ai/blog/introducing-ardus.md): What Ardus does, how it works, and who it is for

## Optional

- [Contact](https://ardus.ai/contact): Map your requirements, or tell us about powered space or spare MW at an existing site
- [Privacy policy](https://ardus.ai/privacy)
- [Terms of service](https://ardus.ai/terms)