Question 1

How fast can a GPU station be provisioned?

Accepted Answer

Configurations held in standing inventory (RTX 5090, RTX 4090 and H100 stations) are typically racked, imaged and handed over within 48 hours. H200 and large clusters depend on current allocation; we give you a firm date on the scoping call.

Question 2

Is the hardware dedicated or shared?

Accepted Answer

Dedicated by default. You get the whole station, root access and isolated networking. Shared on demand instances are available when you only need capacity for a short burst.

Question 3

Can we bring our own images and orchestration?

Accepted Answer

Yes. Most clients run their own containers, schedulers and CI against the stations. We supply a tuned base image with current drivers and CUDA if you would rather start from ours.

Question 4

What happens when a card fails?

Accepted Answer

We monitor the fleet, and replacement is on us. Failed accelerators in leased stations are swapped from local spares. There is no RMA queue for you to manage and no separate support contract.

Question 5

Do you sell hardware outright?

Accepted Answer

We do. Workstations, servers, accelerators and spares are sourced, burned in and shipped with warranty cover. Many clients lease first to validate a configuration, then buy it for steady work.

Question 6

Where is the capacity hosted?

Accepted Answer

In Tier III colocation facilities with redundant power and cooling. We can also deploy into your own facility or cloud tenancy and manage it there.

Privacy policy

What we collect

Why we use it

How long we keep it

Your choices

Cookies