Rent GPU capacity

In this guide:

• Choose how much GPU capacity you need from the listings currently available
• Set an end time and see the cost for that rental window
• Create the account and organization that checkout requires
• Land in Acasia Cloud with the control plane already installed

​

Before you begin

​

Step 1 — Select capacity from the listings

Open the Marketplace and review the available listings. Each listing shows the GPU model, the number of GPUs available, and the supporting CPU, memory, and storage. Select the amount of GPU capacity you want to rent. You are not required to take an entire listing.

​

Step 2 — Set an end time

Choose the date and time the rental should end. The current cost of the rental updates to reflect the window you selected. Adjust the capacity or the end time until the cost and the configuration match what you need.

​

Step 3 — Create your account and organization

Checkout requires a user account and an organization. The organization is the boundary that owns the rental, and it scopes clusters, credentials, users, and billing. If you already have an Acasia account, sign in and select the organization the rental belongs to.

​

Step 4 — Complete payment

Enter your payment details and submit the order. Payment processing is handled by Stripe; Acasia does not store raw card numbers. After payment succeeds:

• A receipt is emailed to the address on the account
• You are signed into Acasia Cloud automatically
• The control plane is installed on the rented capacity without any action from you

​

Step 5 — Start using the capacity

From Acasia Cloud you can create a cluster from the capacity you rented, deploy a model from the model library, and connect over SSH or from your IDE.

​

Summary

1. Select an amount of GPU capacity from the available listings.
2. Pick an end time to see the current cost of the rental.
3. Create a user account and an organization.
4. Complete payment and receive a receipt by email.
5. Land in Acasia Cloud with the control plane installed and start deploying.

​

Next steps