SERVER CONFIGURATOR / CUSTOM DEPLOYMENT

Configure your dedicated GPU server.

Choose GPU model, quantity, runtime, network, memory, and storage. Add deployment details in the notes; the same live API checks inventory and creates the order.

01 GPU & quantity02 System & software03 Network & storage04 Notes & term

PRIMARY WORKLOAD

Primary workload

Choose the main workload so the operator can preconfigure NCCL, drivers, and the container runtime. Click again to clear.

REGION

Choose a region

Estimated network path speed · Edge Probe Reference, for region selection only.

GPU

GPU model and quantity

GPU modelGPU quantity

Standard nodes support up to 8 GPUs; describe higher density in the notes.

SYSTEM & SOFTWARE

Operating system and runtime

Operating systemDriver and runtimeContainers and tools

NETWORK & STORAGE

IP, bandwidth, memory, and storage

Public IPv4IPv6Public bandwidthMemory expansionNVMe storageSSH public keyDeployment notes

BILLING

Term

Rental term