VMware Exit / Cornerstone
VMware exit is a decision framework problem.
Not a vendor selection problem.
The platform that wins your migration depends on six variables — VM count, workload portability, identity model, capex/opex appetite, egress profile, DR requirements, and staff readiness. Pick the framework first, the platform second.
Broadcom's VMware bundles have driven renewal increases of 800–1,500% in mid-market deals. The decision is not whether to evaluate alternatives. The decision is which path fits your workloads.
Mental inputs
What you should know before you scope a migration.
Run these four numbers in your head before you book any vendor calls. The framework on this page assumes you can answer them — at least roughly — within 10 minutes of pulling up vCenter and a renewal quote.
Input 01
VM count
Powered-on VMs across all clusters. Templates and powered-off don't count.
Input 02
Renewal quote
Annual cost on the Broadcom renewal. Compare against the prior-year invoice.
Input 03
On-prem vs cloud preference
Where do you want workloads to live in 5 years? Capex appetite vs OpEx tolerance.
Input 04
Identity model
AD-only, Entra-first, hybrid, or third-party? This drives at least 30% of the recommendation.
The macro driver
Why the renewal math broke.
Broadcom's bundle restructuring after the VMware acquisition consolidated standalone SKUs into VCF (VMware Cloud Foundation) and VVF (VMware vSphere Foundation). For most mid-market customers, that meant being forced into bundles 3–10x larger than the actual usage. Renewal increases of 800–1,500% are common; we have seen 1,400% on a 500-VM estate.
The pricing change is not the entire story. The deeper issue is portfolio risk. When a single vendor controls your hypervisor, your storage virtualization (vSAN), your network virtualization (NSX), and your management plane (vCenter, vRealize), the renewal becomes a take-it-or-leave-it negotiation with no leverage. Customers that diversified before the acquisition closed had options. Customers that did not are now learning what concentration risk costs at renewal time.
The decision framework on this page exists because the question is no longer "should we evaluate alternatives?" That question answered itself the day the renewal quote landed. The real question is which alternatives fit which workloads — and that is a multi-variable problem, not a vendor choice.
The framework
Six decision dimensions.
Score each dimension low / medium / high for your environment. The combination drives the recommendation. No single dimension picks the platform — the matrix does.
Dimension 01
Workload portability
How tightly are workloads coupled to vSphere features? VMs using only standard compute and storage move easily. Workloads leveraging vSAN, NSX micro-segmentation, DRS affinity rules, or vRealize automation need either a destination with parity or rework. Score: low / medium / high coupling.
Dimension 02
Identity tie-in
Is your environment already AD-integrated and Microsoft-stack-heavy? Or are you running multi-tenant, third-party identity? Microsoft-stack shops gain disproportionate value from Hyper-V or Azure (license math + native AD + Azure Arc). Identity-diverse shops should weight Nutanix and Scale Computing more heavily.
Dimension 03
Capex vs opex preference
Hardware refresh cycle and capital appetite drive a lot of this. If your VMware hardware is 4+ years old and CFO is open to OpEx, Azure IaaS gets a serious look. If hardware is recent and capex was budgeted, on-prem destinations (Hyper-V, Nutanix, Azure Stack HCI, Scale Computing) keep the existing investment alive.
Dimension 04
Network egress profile
Workloads with heavy egress (data analytics, file servers, content delivery) make Azure IaaS expensive at scale because egress is metered. Workloads that are mostly chatty east-west traffic don't care. Egress profile alone has flipped Azure-leaning recommendations to Hyper-V or Nutanix multiple times.
Dimension 05
DR / business continuity requirements
What's your RPO/RTO commitment? Hyper-V Replica, Azure Site Recovery, Nutanix Mine, and Scale's snapshot model all hit different price-performance points. Stretched-cluster requirements (sub-second RPO across two data centers) narrow the field hard. Single-site DR opens it back up.
Dimension 06
Staff readiness
Who's running this on day 91? If you have a Microsoft-stack ops team, Hyper-V is operationally cheap because nobody learns a new console. If you have a VMware-trained team and want minimal retraining, Nutanix Prism is closest to vCenter operationally. If you want to minimize ops headcount entirely, Azure IaaS shifts the most operational load to Microsoft.
The matrix
Profile → recommended path.
Six common profiles we see in mid-market VMware estates and the destination that usually wins. Your environment is unlikely to match exactly — the matrix exists so you can locate yourself within 5 minutes and read the rationale.
| Environment profile | Primary recommendation | Why it lands |
|---|---|---|
| Microsoft-stack shop, 100–500 VMs, recent hardware, AD-integrated | Hyper-V on existing hardware | License math (Datacenter includes Hyper-V), native AD, Failover Clustering. Azure Arc bolts on hybrid management without leaving on-prem. |
| Hybrid-curious shop, 250–1,500 VMs, hardware refresh due, OpEx-friendly CFO | Azure Stack HCI + selective Azure IaaS | On-prem performance with Azure operating model. Tier-1 workloads in Azure IaaS, back-office on HCI. One control plane, two locations. |
| vSphere-heavy shop, 200–800 VMs, NSX/vSAN-dependent, VMware-trained team | Nutanix AHV | Operationally closest to vCenter (Prism). Migration tooling (Move) is mature. AHV is included in licensing — no separate hypervisor cost. |
| Cloud-native ambitions, 100–400 VMs, modernization-led, refactor appetite | Azure IaaS + selective PaaS | Treat the migration as the start of modernization. Lift-and-shift Tier-2/3, refactor Tier-1 to App Services + Azure SQL where the business case clears. |
| Operational-simplicity shop, 50–200 VMs, lean IT team, single site | Scale Computing HC3 | Hyperconverged with the simplest operating model in the category. Built-in DR replication. No separate storage team needed. |
| Renewal landed manageable, vSphere deeply integrated, no migration appetite | Stay on VMware | If renewal was 1.5–2x (not 8–15x) and NSX/vSAN are load-bearing, defending the renewal can be the right call. Revisit at next renewal. |
Most real environments split — 60–80% of workloads to a primary destination, the remainder to a secondary that better fits one specific workload class. The matrix is a starting point. The assessment is where the per-workload mapping actually happens.
Walkthrough 01
500-VM manufacturing co — Hyper-V.
Anonymized walkthrough of a recent engagement. The renewal hit was the trigger; the framework picked the destination.
Inputs
- VM count
- 500
- Industry
- Manufacturing (discrete, mid-market)
- Renewal hit
- $310K — up from $22K
- Hardware age
- Year 3 of 5 (Dell servers, recent)
- Identity
- On-prem AD, no cloud identity yet
- Workloads
- ERP, MES, file servers, dev/test, 12 line-of-business apps
Recommendation
Hyper-V on existing hardware
Why: Microsoft-stack shop already running Windows Server Datacenter (Hyper-V is included). Hardware was 3 years old and on the Microsoft HCL. AD-integrated, no cloud identity urgency. Workloads were vanilla — no NSX, no vSAN, no DRS affinity rules. Capex was already on the books for the next two years.
Math: $310K renewal vs $35K migration cost (assessment + project). 4-month payback. Year 2+ run rate dropped to roughly $20K (backup tooling, training).
Timeline: 14 weeks. Discovery and pilot in weeks 1–4. Wave-based cutover weeks 5–12. Two weeks hypercare. VMware decommissioned at week 14.
Walkthrough 02
250-VM SaaS co — Azure IaaS + selective PaaS.
Same framework. Different inputs. Different answer. The point of the framework is that it produces different recommendations for different environments — that's the feature, not a bug.
Inputs
- VM count
- 250
- Industry
- B2B SaaS (sub-100 employee)
- Renewal hit
- $140K — up from $14K
- Hardware age
- Year 5, refresh due in 6 months
- Identity
- Entra ID-first, AD on the way out
- Workloads
- App tier, RDS for partners, dev/test, build agents, observability
Recommendation
Azure IaaS, selective Azure PaaS
Why: Hardware refresh was due in 6 months — perfect window to shift capex to opex. Already Entra-first; AD was a leftover, not a load-bearing system. SaaS workload profile fits cloud-native modeling. The build tier and observability stack moved to PaaS as part of the migration; lift-and-shift was used for the application tier and dev/test.
Math: Hardware refresh avoidance ($180K capex deferred). 3-year Azure run rate roughly $1.2M. The decision wasn't strict TCO — it was strategic. Engineering team wanted to be cloud-native by their next funding round.
Timeline: 16 weeks. Azure landing zone built first (4 weeks). Lift-and-shift waves (8 weeks). PaaS refactor on a parallel track for 3 services. VMware decommissioned at week 16.
What a real assessment delivers
The framework is the entry point. The assessment is the deliverable.
The framework on this page lets you locate yourself in the matrix in 5 minutes. A real engagement produces a written deliverable — workload-by-workload — that the CIO, the CFO, and the audit committee can act on.
Workload inventory + dependency map
Every VM, every dependency, every storage and network path. Surfaces the hidden coupling that breaks naive lift-and-shift estimates.
Per-workload destination recommendation
Not "pick a platform." Per-workload routing — Tier-1 here, back-office there, dev/test somewhere else if the math says so.
5-year TCO model
Including hardware lifecycle, third-party tooling (backup, monitoring, DR), training, and operational run-rate. Not just year-1 license math.
Wave-based migration plan
Cutover sequence, rollback paths, blast-radius management. VMware coexists with the destination through the final wave; no big-bang.
Hypercare definition
Two-week senior-engineer hypercare after final cutover. What's covered, what's not, the on-call rotation, the escalation tree. Plain-English.
Decommission runbook
How VMware comes off the floor. License recovery, hardware reuse or retirement, vCenter sunset, audit documentation. The migration isn't done until the old environment is gone.
Where to go next
Pick a destination read-up.
The framework points at a destination. The destination pages get into the operational specifics — what the migration looks like, what gets retooled, what hypercare covers.
Destination 01
VMware to Hyper-V
Microsoft-stack consolidation using licensing you already own. Failover Clustering, Storage Spaces Direct, native AD.
Read more
Destination 02
VMware to Azure
Azure IaaS lift-and-shift, or Azure Stack HCI for hybrid. Azure Migrate-led, with proper assessment.
Read more
Destination 03
Windows Server EOL
2012/2016/2019 → 2022/2025. In-place upgrade, re-platform, or Azure Arc — picked workload by workload.
Read more
FAQ
Common questions about the framework.
Is VMware exit cheaper than staying on Broadcom?
For mid-market customers in the 100–1,000 VM range, the answer is almost always yes — but not for the reason most procurement teams expect. The renewal-vs-migration math is rarely the deciding factor by year three, because the new platform's run-rate cost catches up quickly. The real cost driver is the 5-year TCO including hardware lifecycle, staff retraining, third-party tooling (backup, monitoring, DR), and the operational debt of running a stack you no longer want to be on. Run that math and Broadcom usually loses, but the winner depends entirely on workload mix.
How do you decide between Hyper-V, Azure, Nutanix, and Azure Stack HCI?
We don't pick one platform and migrate everything to it. We assess workloads against six dimensions (portability, identity tie-in, capex/opex preference, network egress, DR requirements, staff readiness) and recommend per-workload. Most mid-market estates split: 60–80% to one primary destination, the remainder to a secondary platform that better fits a specific workload class. Tier-1 SaaS-style applications often go to Azure IaaS while back-office workloads land on Hyper-V or Nutanix.
We have 250 VMs. Is that mid-market or do we need a Tier-1 firm?
250 VMs is squarely mid-market and squarely in our wheelhouse. Tier-1 firms (Accenture, Deloitte, the global SIs) typically minimum-quote at engagements 5–10x larger than that. Commodity MSPs that quote at this size usually don't have senior virtualization engineers on staff — they hand the work to whoever is available. The work at this size is too large for a one-person consultant and too small for a Tier-1 firm. That gap is what we built Pro IT NW to fill.
What if we want to stay on VMware?
Sometimes that's the right call. If your renewal landed at a manageable increase, your workloads are deeply integrated with NSX or vSAN, and you have no appetite for a 12-month migration project, staying is a defensible decision. The decision framework still applies — you're answering the same six questions, just arriving at a different recommendation. We've told customers to renew with Broadcom and we'll tell you the same if the math says so.
Is this a sales call disguised as a framework?
The framework on this page is the actual framework. We're not holding back the answers to drive a phone call. The scoping conversation exists because every environment has nuances the framework doesn't capture — workload dependencies, application-vendor support matrices, contract timing, internal political dynamics. The 30-minute scoping call is free; the assessment that follows is fixed-fee.
Field notes
- → VMware to Azure vs Hyper-V: 500-VM decision guide — the head-to-head decision on the two destinations that win most mid-market migrations, with two anonymized walk-throughs scored across all six dimensions.
- → After Broadcom: 250-VM Hyper-V migration in 90 days — the field-notes timeline of a real Broadcom-renewal exit at 250-VM scale.
- → VMware after Broadcom: your real exit options — the long-form deep-dive on Hyper-V vs Nutanix vs Azure Stack HCI vs Scale Computing vs Proxmox vs Azure IaaS, with the math on when each one wins.
Run your environment through the framework.
Tell us the VM count, the renewal date, and one sentence about your identity model. Two-business-day response with a 30-minute scoping call slot.