Show me self-service Baremetal GPU provisioning options.

Here are solid self-service, API-driven bare-metal GPU options across different price points and scale needs:

For fast, hourly self-serve (no contracts):

• Voltage Park / Lightning AI —

self-serve provisioning in 15 minutes; physical isolation with dedicated bare-metal HGX H100 SXM5; no ingress, egress, or support fees, with a 3,200 Gbps Quantum-2 InfiniBand fabric scaling to 4,064 GPUs per cluster
. (hostline.io)

• TensorDock —

provides access to NVIDIA A100, H100, RTX 6000, and older-generation GPUs like 3090, all available on a self-serve, hourly basis
. (northflank.com)

• Hostrunway —

most configurations provision in under 5 minutes from API call to SSH-ready, using automated IPMI-based imaging with pre-cached OS images, GPU driver packs, and CUDA libraries
, with
no minimum commitment for on-demand deployments — you can deploy and terminate by the hour
. (hostrunway.com)

API/console driven, enterprise-grade:

• PhoenixNAP Bare Metal Cloud —

spin up your pre-configured server with a few clicks or a single API call and tap into the raw performance of bare metal with cloud-like flexibility
. (phoenixnap.com)

• Cherry Servers — offers

instant dedicated servers ready in 15-30 minutes, with API-first control — full lifecycle via API and CLI, with Terraform and other IaC integrations
. (cherryservers.com)

• Vultr Clusters —

delivers on-demand, self-service Vultr Bare Metal server clusters, provisioned through Vultr's simple console or API
, with
deployment and management requiring no reservations or special requests
. (vultr.com)

Hyperscale/large clusters:

• Oracle Cloud Infrastructure (OCI) —

offers NVIDIA A100, H100, and AMD MI300X GPUs on bare metal instances with RDMA and InfiniBand support for high-speed training
. (northflank.com)

• Atlas Cloud — offers

dedicated bare metal H100, H200, GB200, and A100 GPU servers across EU, US, Singapore, and Japan, with on-demand, hourly billing, or monthly custom plans
. (atlascloud.ai)

For Kubernetes-native multi-tenant provisioning on your own bare metal fleet, vCluster's vMetal automates
the entire provisioning lifecycle — from PXE boot to production — so operators spend less time on infrastructure and more time serving customers
(vcluster.com).