Enterprise Virtualization, Without the Per-Feature Licensing

Run Your Own
Hypervisor Cloud

MFCloud is a multi-tenant KVM control plane: VM lifecycle, live migration, HA failover,
GPU passthrough, and Ceph/ZFS/LINSTOR storage — one console, your own hardware.

Rocky Linux • KVM • libvirt • Open vSwitch / OVN • Ceph • ZFS • LINSTOR/DRBD

Get Started Free Read the Docs

Hardware Acceleration

GPU Passthrough
Included, Not an Add-On

Attach a physical GPU directly to a VM over PCIe/VFIO — full driver access, near-native performance, no separate license tier.

PCIe / VFIO Passthrough

Bind any IOMMU-isolated GPU to a node and attach it to a VM from the deploy dialog — Q35 + UEFI wired up automatically so large VRAM BARs map correctly.

Works with Windows and Linux guests
Full Resizable BAR support for modern GPUs
Hardware passthrough picker built into Deploy VM

No Per-GPU License Fee

Other virtualization platforms gate GPU passthrough and vGPU behind separate paid add-ons. In MFCloud it's a standard platform capability.

Included on the free tier
No "GPU edition" upsell
Bring your own supported GPU hardware

The Platform

Everything an Operator Needs

One console for VM lifecycle, storage, networking, identity, and HA — across as many nodes as you run.

VM Lifecycle

Deploy, snapshot, clone, live-migrate, and template VMs across your fleet

HA & Failover

Automatic fencing + VM restart when a node goes down

Multi-Tenant RBAC

Super Admin, Tenant Admin, Operator & Viewer roles with per-tenant quotas

V2V Import

Bring existing VMs in from OVA/OVF/VMX/disk images — straight to KVM

Backup & DR

Datastore snapshots and agentless encrypted backup to S3-compatible storage

Software-Defined Networking

OVN logical switches/routers, security groups, and IPAM pools

Containers Too

LXD/Incus containers alongside VMs, on the same nodes

LDAP / SSO

Active Directory, OIDC, Entra ID, Google Workspace, and Okta sign-in

Bring Your Own Storage

Storage Backends That Fit
However You Run Today

Pick the backend per cluster — MFCloud doesn't force one storage architecture on your whole fleet.

Ceph RBD

Wide multi-host pools, snapshots, scale-out clusters

LINSTOR / DRBD

Replicated block storage tuned for HA, low latency

ZFS

Single-host pools, snapshots, and scratch volumes

NFS

Point at storage you already have on the network

See It In Action

One Console,
Your Whole Stack

VMs, storage, networking, Kubernetes and live telemetry — managed from a single pane of glass, running on your own hardware.

MFConsole unified dashboard — cluster capacity, control plane, and built-in AI assistant — **Unified dashboard** — nodes, VMs, and resource usage at a glance

Virtual machine management — HA-protected guests running on a Ceph-backed cluster — **VM management** — HA-protected guests on shared Ceph storage

Live performance telemetry — CPU, RAM, IOWAIT and VM counts with real-time charts — **Performance & monitoring** — real-time metrics across the fleet

Software-defined networking — Open vSwitch portgroups and cluster VLAN fabric — **Software-defined networking** — OVS portgroups & per-cluster VLAN fabric

OVN overlay tunnel endpoints, L2 bridges, and IP address management — **OVN overlay & IPAM** — Geneve tunnels, L2 bridges & address pools

Hyperconverged Ceph storage — cluster health, capacity, and active OSDs across the fleet — **Hyperconverged Ceph storage** — health, capacity & OSDs, managed in-console

Run Your Own Hypervisor Cloud

GPU PassthroughIncluded, Not an Add-On