New: OpenVZ containers now shipping on every NVMe tier.See plans →
Virtual machine hosting

Pick your hypervisor.
Root your own box.

SiteServing runs XEN, KVM, and OpenVZ side by side on the same AMD EPYC / NVMe fleet, so you can deploy a VM on whichever virtualization model your stack actually needs, with full root access from the first boot.

99.9% uptime SLA DDoS filtering included Free migration, zero downtime
Virtualization

One fleet, three virtualization models

Every hypervisor sits on identical hardware. The difference is how much of the kernel your VM gets to itself, so pick the model that matches your workload, not the other way around.

Dom0 + DomU

XEN

A thin hypervisor runs directly on the hardware. A privileged control domain (Dom0) manages a set of guest domains (DomU), each running a hypervisor-aware kernel through paravirtualized drivers, delivering near-native performance with hardware-level separation between guests.

  • IsolationHardware-level, separate kernel
  • Custom kernelYes
  • Typical overheadLow
  • Live migrationYes
  • Best forCustom kernels, GPU passthrough, strict isolation
hardware: amd epyc / nvme xen hypervisor (thin) dom0 control domU guest kernel domU guest kernel domU guest kernel
Kernel-based VM

KVM

KVM turns the Linux kernel itself into a hypervisor using hardware virtualization extensions (Intel VT-x / AMD-V). Each guest runs a complete, independent kernel and behaves like it's on dedicated hardware, including unmodified Windows or BSD images.

  • IsolationHardware-level, separate kernel
  • Custom kernelYes, any OS
  • Typical overheadLow–moderate
  • Live migrationYes
  • Best forUnmodified OS images, Windows guests, general purpose
hardware: amd epyc / nvme linux kernel + kvm module guest vm full kernel guest vm full kernel guest vm full kernel
Namespace isolation

OpenVZ

Containers share a single host kernel and isolate everything else (filesystem, processes, network) through kernel namespaces. No per-guest hypervisor overhead means more density per node, at the cost of running the host's kernel version.

  • IsolationProcess / namespace-level
  • Custom kernelNo, shares host kernel
  • Typical overheadMinimal
  • Live migrationLimited
  • Best forHigh-density budget VPS, predictable Linux workloads
hardware: amd epyc / nvme single shared host kernel container container container container
VPS hosting

Sized like server specs should be

Every plan runs on AMD EPYC hosts with DDR5 memory and NVMe storage. Choose XEN, KVM, or OpenVZ at checkout: same hardware, same price, root access on all three.

Node.S
$6/mobilled annually · renews $13/mo
  • vCPU1 core
  • Memory2 GB DDR5
  • Storage40 GB NVMe
  • Transfer2 TB
xenkvmopenvz
Deploy Node.S
Node.L
$24/mobilled annually · renews $49/mo
  • vCPU4 cores
  • Memory8 GB DDR5
  • Storage160 GB NVMe
  • Transfer8 TB
xenkvm
Deploy Node.L
Node.XL
$48/mobilled annually · renews $95/mo
  • vCPU8 cores
  • Memory16 GB DDR5
  • Storage320 GB NVMe
  • TransferUnmetered
xenkvm
Deploy Node.XL

Every plan includes a dedicated IP, root SSH and API access, network-level DDoS filtering, and free migration from your current host within 30 days of signup.

Infrastructure

Built to stay up

Every node is monitored around the clock across redundant upstream networks, with DDoS filtering applied before traffic ever reaches your VM.

99.9%
uptime SLA
24/7
infrastructure monitoring
6
redundant upstream providers
<15m
avg. support first response
US-East · IBX DC11
online · 8ms
US-West · IBX LA04
online · 11ms
EU-Central · IBX DU01
online · 14ms