VMware Broadcom Transition Guide 2026
How Broadcom rebuilt VMware's commercial model around subscription bundles, and the field-tested moves enterprise buyers use to contain a 2 to 5x cost shock, preserve use, and keep a credible exit on the table.
You are registered. Your guide is ready. Read the full 2026 edition of the VMware Broadcom Transition Guide below.
Prepared by Atonement Licensing · buyer-side advisory · last reviewed June 2026. List rates, per-core figures, and the worked cluster example below are clearly labelled indicative benchmarks for illustration, not quotes; Broadcom's bundle pricing is negotiated and varies by account, region, and term.
Executive summary
Broadcom's first VMware renewal quote is an anchor built on your perceived lock-in, and the single most powerful lever for bringing it down is a credible, costed alternative. Since closing its acquisition of VMware in November 2023, Broadcom has executed one of the most aggressive commercial transformations in enterprise software history: perpetual licences were discontinued and replaced with subscription-only terms, the catalogue was collapsed from dozens of SKUs into a small number of bundles led by VMware Cloud Foundation (VCF) and VMware vSphere Foundation (VVF), Support and Subscription was folded into the subscription, and minimum purchase thresholds were raised, first to 16 cores per CPU and, for many customers, toward much larger account-wide minimums.
The result for a typical enterprise renewal is a quoted increase of two to five times prior spend, and in outlier cases ten times or more. On an indicative eight-host cluster that previously cost on the order of $60,000 a year in perpetual support, the same estate re-quoted under VCF per-core subscription terms models toward $250,000 to $300,000 a year — a four-to-five-fold step on identical hardware. Yet the gap between Broadcom's opening number and a defensible one is wide: edition discipline, core-count rigour, and a believable exit routinely recover 20 to 50% of the proposal.
This guide distils what former VMware and Broadcom commercial practitioners understand about how the new model is engineered and where it can be challenged: the bundle economics, the per-core subscription mechanics and the minimum-commitment trap, the perpetual-to-subscription cliff, the realistic alternatives, a structured negotiation framework, and a migration timeline that makes an exit credible whether or not you ever execute it. Read it before you respond to a quote, not after.
What Broadcom actually changed
The transition is not a price rise layered on the old model, it is a different commercial architecture, and each structural change is designed to convert what was previously optional spend into recurring, bundled, committed spend. Three changes do most of the work.
First, perpetual licensing ended. Existing perpetual licences remain valid, but they can no longer be renewed for support; once Support and Subscription lapses, the only path to patches, security fixes, and upgrades is a subscription. This removes the "run it as-is for another three years" option that many infrastructure teams relied on as a fallback. Second, the catalogue was consolidated into bundles. Where customers once bought vSphere standalone and added vSAN, NSX, or Aria selectively, the headline offerings are now VMware Cloud Foundation, the full stack of compute, storage, networking, and management, and the lighter vSphere Foundation. Customers who used only a fraction of the stack are steered toward paying for the whole bundle. Third, pricing moved to a per-core subscription with rising minimums, so even lightly populated CPUs are billed at a floor that often exceeds prior usage.
Understanding this architecture matters because each lever is negotiated differently. Bundle composition is contestable through edition selection; per-core cost is contestable through core-count discipline and term; and the migration cliff is contestable only by having somewhere else to go.
Broadcom's commercial teams are measured on total contract value and multi-year commitment, not on logo retention at any price. That is why the opening quote is deliberately high, and why the deepest concessions appear late in the cycle and almost always require evidence that you are seriously evaluating an alternative. The number moves when your alternative becomes believable, not when you object to the number.
Action. Map your three levers before you reply to anything: which bundle components you actually run, how many cores you truly need licensed, and what your exit would cost. Each is a separate negotiation.
2The bundle economics: VCF versus VVF
The two anchor products define the decision space. VMware Cloud Foundation (VCF) is the premium full-stack bundle and carries the higher per-core rate; VMware vSphere Foundation (VVF) is the lighter bundle aimed at customers who need virtualisation and management but not the integrated storage and networking layers. Broadcom's sales motion strongly favours steering customers to VCF, because it captures the largest share of wallet and locks in the most components. For a buyer, the first analytical task is brutally honest: which components of the stack are you actually running in production, and which are being attached because the bundle makes them appear free?
The table below sketches the indicative positioning of the tiers. List rates move and are negotiable; the point is the relative gap and what each tier includes.
| Bundle | Core components | Indicative list / core / year | Best fit |
|---|---|---|---|
| vSphere Standard (where offered) | vSphere only, capped | ~$50 | Small clusters, edge, ROBO sites |
| vSphere Foundation (VVF) | vSphere, vCenter, Aria management, limited vSAN capacity | ~$135 | Compute-centric estates using third-party storage and networking |
| Cloud Foundation (VCF) | vSphere, vSAN, NSX, Aria suite, full SDDC stack | ~$350 | Private-cloud adopters running the full software-defined data centre |
The single most common overspend we see is a customer accepting VCF when their production reality is "vSphere plus a bit of vSAN." If NSX and the full Aria suite are not deployed and not on a near-term roadmap, paying the VCF premium for them is pure leakage. The bundle's extras are only a saving if you would otherwise have bought them.
Action. Inventory what is genuinely in production — vSAN, NSX, Aria — and price VVF against VCF on that footprint. Make Broadcom justify the premium component by component.
3Per-core mechanics and the minimum-commitment trap
Broadcom prices per physical core, with a minimum of 16 cores billed per CPU regardless of the actual core count. For older or smaller processors this floor inflates the bill; for modern high-core-count CPUs the minimum is less of a factor but the per-core rate dominates. Critically, many enterprise agreements now also carry an overall minimum core commitment for the account, which can force customers to licence well above their currently deployed footprint.
The arithmetic compounds quickly. A modest cluster that cost a few tens of thousands of dollars per year in perpetual support can re-quote into the high six figures once every core is subscription-licensed at bundle rates. The worked example below shows how the same eight-host cluster lands under the old and new models, using clearly indicative figures.
| Line | Perpetual + SnS (old) | VCF per-core subscription (new) |
|---|---|---|
| Hosts × CPUs × cores | 8 hosts, 2 CPUs, 32 cores each | 8 hosts, 2 CPUs, 32 cores each |
| Licensable cores | n/a (perpetual, paid once) | 512 cores billed annually |
| Indicative annual cost | ~$60,000 SnS | ~$250,000 to $300,000 |
| Step change | baseline | ~4 to 5x (indicative) |
Two disciplines blunt the per-core mechanics. The first is core-count rigour: licence the cores you run, resist account-wide minimums that exceed your footprint, and consolidate workloads onto fewer, denser hosts before a renewal so you are not paying a 16-core floor on half-empty sockets. The second is term: a multi-year commitment earns a deeper rate, but only commit long when your footprint is stable and the deal carries a price hold, because a long lock on an inflated core count is the most expensive mistake in this market.
Action. Build your own core-count model from the live estate, consolidate where you can, and challenge any account minimum that licences cores you do not deploy.
4The perpetual-to-subscription cliff
The cliff is the moment support lapses on a perpetual estate and the only forward path is a subscription. It is engineered pressure: your licences still run, but unpatched infrastructure is a security and compliance exposure most enterprises cannot carry for long, so the practical deadline is tighter than the legal one. Broadcom's leverage at the renewal table rests almost entirely on how close you are to that edge and how few options you appear to have.
Buyers blunt the cliff by separating the support question from the platform question. Running a stable perpetual estate unsupported for a defined, risk-assessed window — or bridging it with third-party support where appropriate — buys negotiating time and breaks the false urgency. The goal is not to run unsupported indefinitely; it is to convert a "sign now or go dark" ultimatum into a planned decision made on your calendar.
The range we see recovered from a first VCF proposal through edition discipline, core-count rigour, and a credible alternative, before any goodwill concession (indicative).
The realistic time to stand up and migrate to an alternative platform — long enough that the exit must be started early to be credible at the table (indicative).
Action. Risk-assess a defined unsupported or third-party-supported window for your most stable workloads, so a lapsing support date stops being Broadcom's deadline and becomes your planning horizon.
5The realistic alternatives
An alternative does not have to be deployed to work in a negotiation; it has to be credible, costed, and visible. The hypervisor and private-cloud market has more genuine substitutes than it did two years ago, and Broadcom's own pricing has done more to make them attractive than any competitor's marketing. The point of the table below is not that any one option is right, but that each is real enough to cost and to put on the table.
| Alternative | Best fit | Main switching cost |
|---|---|---|
| Nutanix (AHV) | HCI estates wanting an integrated stack and a managed migration path | Re-platforming, retraining, hardware compatibility |
| Microsoft Azure Local | Microsoft-centric estates with hybrid and Azure Arc alignment | Operational model change, Microsoft licensing interplay |
| Proxmox VE / open-source KVM | Cost-sensitive or technically deep teams comfortable owning support | In-house operational maturity, support model |
| Public cloud (AWS, Azure, GCP) | Workloads suited to refactor or lift-and-shift away from on-prem | Re-architecture, run-rate economics, egress |
Migration is real work — 12 to 24 months for a full estate is a fair planning assumption — which is precisely why an exit has to be started early to be believable when it matters. A proof of concept on a non-critical workload, a costed target architecture, and a named internal owner convert "we might look at Nutanix" into a fact Broadcom's deal desk has to price against.
Broadcom's number moves when your alternative becomes believable, not when you tell them the price is too high.
Action. Stand up a small, real evaluation of at least one alternative on a non-critical workload before your renewal window, and let the account team know it exists.
Facing a Broadcom renewal or a 2 to 5x quote? Our advisors model your estate and run the negotiation with you.
Cloud Contract NegotiationThe negotiation framework
A Broadcom negotiation is won on three fronts at once: the bundle, the core count, and the alternative. Edition discipline strips the components you do not run; core-count rigour challenges minimums that exceed your footprint; and a credible exit supplies the leverage that turns objections into concessions. Run them separately and the account team picks you off on each; run them together and the proposal has nowhere to hide.
Benchmark every element. The per-core rate, the bundle premium, the multi-year discount, and any one-time migration or transition funds should each be normalised and compared against what comparable enterprises secure, not against Broadcom's own list. Insist that concessions are expressed as annual cost per core for your actual footprint over the full term, with one-time funds stripped out and shown separately, so a "generous" headline does not disguise a weak structure. And escalate: the account team owns the relationship, but the deepest exception pricing is approved above them and appears late, almost always once an alternative is visible.
Action. Build a single normalised comparison — opening quote, your counter, and your benchmarked target — and make Broadcom respond to your structure rather than defending theirs.
7Building a credible exit: the migration timeline
The exit that protects your renewal is the one you began before you needed it. A migration started the month a quote lands is not credible; one with a proof of concept already running and a costed target architecture is. The three phases below are how we sequence it so the option is real at the table even if you ultimately re-sign with Broadcom on better terms.
Assess and prove
Inventory the live estate, identify which workloads are portable, and stand up a proof of concept on one non-critical cluster with a costed target architecture.
Pilot and cost
Migrate a contained workload in earnest, capture the true switching cost and run rate, and build the business case that makes the alternative defensible to your board and visible to Broadcom.
Decide and execute
Take the renewal with a real exit in hand: re-sign on improved terms, migrate in waves, or run a hybrid estate. The leverage holds either way.
Action. Put the migration assessment on the calendar at least 12 months before your support lapse, owned by a named infrastructure lead, so the exit is a fact and not a threat.
Treat Broadcom's first quote as an anchor, not a price. Right-size the bundle to deployed reality, hold the core count to your real footprint, separate the support deadline from the platform decision, and start a credible alternative early enough that it is real at the table. Normalise every concession to annual cost per core over the term, escalate for exception pricing once your exit is visible, and time the close to a Broadcom quarter end. The 20 to 50% that separates the opening number from a defensible one is recovered by structure and leverage, not by objection.
Key takeaways
- The Broadcom model is a different commercial architecture, not a price rise: perpetual ended, the catalogue became bundles, and pricing moved to per-core subscription with rising minimums.
- Typical renewals quote 2 to 5x prior spend; the gap to a defensible number is wide and recovered through edition discipline, core-count rigour, and a credible exit.
- Decide VCF versus VVF on deployed reality — paying the VCF premium for NSX and Aria you do not run is pure leakage.
- Challenge the 16-core-per-CPU floor and any account-wide minimum that licences cores beyond your footprint; consolidate before you renew.
- Separate the support cliff from the platform decision to break false urgency and buy negotiating time.
- Nutanix, Azure Local, Proxmox, and public cloud are real alternatives; a costed proof of concept is the single strongest lever on the price.
- Start the migration assessment 12 to 24 months out so the exit is credible whether or not you execute it.
Frequently asked questions
How much do VMware costs increase under Broadcom?
Most enterprise renewals are quoted at two to five times prior spend, driven by subscription-only terms, bundle consolidation, and per-core minimums, with outlier cases of ten times or more. The increase is an opening anchor, not a fixed outcome; edition discipline, core-count rigour, and a credible alternative routinely recover 20 to 50% of the proposal.
Can we keep our existing perpetual VMware licences?
Yes, existing perpetual licences remain valid and keep running, but they can no longer be renewed for support. Once Support and Subscription lapses, the only path to patches, security fixes, and upgrades is a subscription, which is the pressure point Broadcom relies on at renewal.
Should we choose VCF or VVF?
Decide on deployed reality. VMware Cloud Foundation is the full stack and the higher per-core rate; vSphere Foundation is the lighter bundle for compute-centric estates using third-party storage and networking. If NSX and the full Aria suite are not in production and not on a near-term roadmap, VVF plus targeted add-ons usually beats paying the VCF premium.
What is the 16-core minimum?
Broadcom bills a minimum of 16 cores per CPU regardless of the actual core count, and many agreements also carry an account-wide minimum core commitment. Both can force you to licence above your deployed footprint, which is why core-count rigour and workload consolidation before a renewal matter as much as the per-core rate.
What are the realistic alternatives to VMware?
Nutanix, Microsoft Azure Local, Proxmox and open-source KVM, and public cloud are all genuine options depending on the workload. Full migration is a 12-to-24-month effort, so an alternative has to be started early to be credible — but a costed proof of concept on a non-critical workload is the strongest single lever on Broadcom's price, whether or not you ultimately move.
Get this guide applied to your Broadcom renewal. Confidential assessment within one business day.
Book a 30 minute callRelated research: the VMware Broadcom Renewal Defense playbook, the cloud renewal strategy playbook, and the Microsoft Enterprise Agreement Guide for the parallel subscription-bundle playbook.