For teams leaving Broadcom-era VMware but who still want a turnkey, vendor-supported platform rather than self-managed open source, Nutanix AHV is the usual destination. AHV is Nutanix’s KVM-based hypervisor, and crucially it is included at no extra license cost with the Nutanix Cloud Infrastructure (AOS) stack, you pay for the platform, not separately for the hypervisor. This is a different trade-off from Proxmox or XCP-ng: you’re swapping one commercial vendor for another, but escaping per-core hypervisor subscriptions and getting integrated hyperconverged storage and management.
What you gain, and what it costs
You gain a tightly integrated HCI platform: compute, distributed storage, and management in one stack, with first-party 24×7 support and a polished UI. AHV itself carries no per-core hypervisor fee. You give up the option of running on arbitrary commodity hardware, Nutanix runs on Nutanix appliances or validated nodes, and you take on AOS licensing (per-node/core subscription). Put plainly: this is the move for organizations that valued vSphere’s turnkey support and want the same experience, not the maximum-savings open-source route.
Architecture mapping
- vCenter → Prism Central / Prism Element. Prism is the management plane across clusters.
- vSAN → Nutanix distributed storage (AOS). Storage is built into the platform; no separate SAN.
- vSphere HA → built-in AHV HA, configured per VM.
- DRS → Acropolis Dynamic Scheduling (ADS), automated workload balancing.
- Distributed vSwitch → AHV networking (managed/unmanaged networks, VLANs).
- vSphere snapshots / SRM → Nutanix snapshots, Protection Domains, and async/sync replication for DR.
The V2V, done with Nutanix Move
The standout tool is Nutanix Move, a free, agentless migration appliance. You connect it to vCenter, select VMs, and it replicates them to AHV while they keep running, then performs a short final cutover sync, keeping downtime to minutes. Move also handles much of the guest prep automatically.
Post-migration steps that still matter: install Nutanix Guest Tools (NGT) so the platform integrates with the guest (and to inject VirtIO drivers on Windows), verify boot (UEFI vs BIOS preserved), and re-check NIC/IP. Windows guests need the VirtIO drivers present before relying on the virtio storage path, NGT handles this, but validate on a pilot VM.
Move runs in successive passes: an initial seed replication copies the bulk of each disk while the source keeps serving traffic, then incremental passes keep the AHV copy current until you schedule the cutover. Because the heavy copy happens ahead of time, the actual outage is confined to the final sync and a reboot on AHV, which is why teams often batch cutovers into a single maintenance evening. Order the batches so shared services and dependencies migrate before the workloads that lean on them, and stage the replication of large or busy VMs early so their seed pass has time to complete. Watch the source’s change rate during seeding: a very write-heavy VM lengthens each incremental pass, so give those extra runway before cutover.
Size on nodes, not just cores
Nutanix sizing is node- and cluster-oriented: you size for compute, storage capacity, and the resiliency factor (RF2/RF3 replication consumes raw capacity). Budget for the replication overhead and N+1 node redundancy. The cost comparison is “vSphere per-core subscription + separate SAN” versus “Nutanix platform subscription with storage included”, model both honestly, since Nutanix licensing is real even though AHV is free.
What you rebuild rather than migrate
Even with Move automating the bulk of the work, some vSphere constructs need a rebuild rather than a migration:
- vSAN storage policies. Resilience on Nutanix comes from the cluster’s replication factor (RF2/RF3), not per-VM SPBM. Translate the intent of each policy into the cluster design rather than expecting a mapping.
- NSX overlays and distributed firewall. Recreate segmentation with AHV managed/unmanaged networks and VLANs, plus Flow microsegmentation if licensed; there is no NSX import.
- Fault Tolerance. AHV provides HA restart, not lockstep FT, so continuous-availability workloads need application-level clustering.
- SRM runbooks. DR moves to Nutanix Protection Domains with async or sync replication, which means rebuilding recovery plans rather than importing SRM.
Per-VM sign-off
Confirm the same items on each migrated VM so the runbook stays consistent:
- The guest boots on the preserved firmware (UEFI vs BIOS) to a login prompt.
- Nutanix Guest Tools (NGT) is installed and the VirtIO storage path is in use on Windows.
- The NIC is on the intended AHV network and VLAN with the right static IP, gateway, and DNS.
- VMware Tools has been removed so no stale drivers remain.
- A Nutanix snapshot is taken and a restore boots successfully.
- AHV HA is enabled where required and a simulated node failure restarts the VM.
Staging the rollout
Stand up the Nutanix cluster, deploy Move, and pilot a few low-risk VMs end to end: replicate → install NGT → validate → snapshot/restore test → HA failover. Capture the runbook, then migrate wave by wave. Keep the source VMs intact on vSphere through a hypercare window for instant rollback.
Who this move is for
VMware vSphere → Nutanix AHV suits teams that want to keep a supported, turnkey platform while shedding Broadcom’s per-core hypervisor pricing, AHV is free, Move makes the V2V low-downtime, and storage/DR are built in. The trade-offs are validated hardware and AOS licensing. Model “vSphere + SAN” vs “Nutanix platform” in the calculator above, and treat the figures as illustrative until Nutanix or a partner quotes your node configuration.