Broadcom moved VMware to subscription only and reset what a virtualization estate costs. This report reads the transition reset from real renewals, maps the bundles, and shows how prepared buyers cap the increase.
The Broadcom VMware subscription reset is the steepest enterprise software price move we tracked in 2026, and this report reads what it actually costs a perpetual estate and how prepared buyers cap it.
About this report. The figures here are defensible bands, not point estimates. They come from the Redress Compliance advisory engagement file for 2024 and 2025, cross checked against public Broadcom and VMware sources.
We benchmarked transitions across estate sizes, regions, and industries. Where a number is a single client outcome we say so. Everywhere else, treat the band as a planning range, not a quote for your estate.
The headline answer is that a perpetual VMware estate converting to the new subscription bundles typically sees its annual cost reset by more than 100 percent. That is the transition reset, and it is a one time repricing, not an annual uplift.
After the reset, the recurring subscription still carries an annual uplift at renewal, usually in the high single digits to low teens. Gartner has reported that software is the fastest growing slice of enterprise IT spend, and the VMware move is a sharp example.
The transition reset is the gap between your old perpetual plus maintenance cost and the new subscription cost for the same capacity. The annual uplift is the smaller percentage applied to the subscription each year after that.
Buyers who confuse the two negotiate the wrong number. The reset is a one time event you fight once. The uplift is a recurring term you cap for the life of the deal.
Public cases sit at the extreme end of the band. One large carrier disclosed a quote it described as a 1,050 percent increase, and several estates reported multiples rather than percentages. These are outliers, not the median, but they show the ceiling.
Read the reset against your fully loaded prior cost, not the list price of one product. Include perpetual amortization, maintenance, and any add ons. The honest comparison is total annual cost before versus total annual cost after.
Broadcom replaced a catalog of more than 8,000 SKUs with a short list of subscription bundles to simplify selling and raise the average deal size. The headline bundles are VMware Cloud Foundation and vSphere Foundation, a shift Broadcom set out in its company newsroom.
Broadcom positions VMware Cloud Foundation as the flagship full stack, sold per core on a subscription. Lower tiers built on vSphere cover smaller estates, but the entry point is higher than the old per socket perpetual world.
The Broadcom VMware bundle map for 2026, simplified
| Bundle | What it includes | Sold by | Typical fit |
|---|---|---|---|
| VMware Cloud Foundation | Full private cloud: vSphere, vSAN, NSX, Aria | Per core subscription | Large estates standardizing on a private cloud |
| vSphere Foundation | vSphere plus vSAN entry and Aria operations | Per core subscription | Mid estates wanting more than compute |
| vSphere Enterprise Plus | Compute virtualization, advanced features | Per core subscription | Estates that only need vSphere |
| vSphere Standard | Compute virtualization, reduced features | Per core subscription | Small estates, limited availability |
vSphere Standard as a freely available standalone perpetual product was discontinued. It returned in limited subscription form, but the cheap entry that small estates relied on is gone. Many are now quoted a richer bundle than they need.
Most estates are quoted VMware Cloud Foundation first because it carries the highest per core price. The buyer side question is whether vSphere Foundation or vSphere Enterprise Plus covers the actual workload at a lower cost.
The old VMware model often priced per processor or per socket. Broadcom prices per core, with a minimum of sixteen cores per processor. Dense modern processors carry far more than sixteen cores, so the model raises cost on exactly the hardware buyers run today.
The consolidation removed the ability to buy only the piece you need. Standalone products folded into bundles, so a buyer who used one feature now pays for the suite. Simplicity for the seller became a higher floor for the buyer.
The shift converts a fully amortized perpetual license, which carried only a maintenance fee of roughly 20 to 25 percent a year, into a full recurring subscription. For a paid off estate, the annual cash cost can more than double.
This is the heart of why the reset feels so steep. Buyers were paying maintenance on an asset they already owned. Now they pay a license fee every year, indefinitely, with no equity building up.
On a perpetual license the annual maintenance fee typically ran 20 to 25 percent of the original license price. The subscription replaces that with a full annual license fee, so the same capacity can cost three to five times the old maintenance line.
Over three years the subscription removes the comfort of a paid off asset. Instead of maintenance on owned licenses, you carry a full fee each year. The cumulative cash difference is what makes the reset a board level topic.
Illustrative three year cash effect for a paid off estate, indexed
| Year | Old perpetual plus maintenance | New subscription | Difference |
|---|---|---|---|
| Year 1 | 24 | 100 | plus 76 |
| Year 2 | 24 | 108 | plus 84 |
| Year 3 | 24 | 117 | plus 93 |
No. Broadcom ended new perpetual VMware sales in 2024 and has not signaled a reversal. Existing perpetual licenses can run, but they lose support over time, so the practical horizon is a planned migration or a subscription conversion.
The credible alternatives are Nutanix, Microsoft Hyper V, Proxmox, public cloud, and an OpenStack or KVM build for the largest estates. Each trades license savings against migration cost and operational change.
Each alternative suits a different estate. The right choice depends on your workload mix, your existing vendor agreements, and how much operational change your team can absorb.
A migration pays off when the multi year license saving clears the one time migration cost and the added operational risk. In our engagements the break even on a credible alternative typically sits 9 to 14 months out, longer for storage heavy estates.
When a VMware migration pays off, by workload profile
| Workload profile | Migration effort | Break even window | Best fit alternative |
|---|---|---|---|
| Compute heavy, simple storage | Moderate | 9 to 12 months | Microsoft Hyper V or Proxmox |
| Mixed, vSAN dependent | High | 14 to 24 months | Nutanix |
| Cloud ready, modernizing | High up front | Varies | Public cloud rehost |
| Small, non production | Low | 6 to 9 months | Proxmox or KVM |
A migration is more than swapping a hypervisor. It touches storage, networking, backup, monitoring, and the operating model your team runs every day. The license saving is real, but the project cost and the learning curve are where migrations slip.
The hidden costs are dual running, retraining, and temporary tooling during the cutover. Most estates run both platforms for a period, which adds cost before it removes it. Budget the overlap honestly or the break even slips.
Buyers cut the Broadcom increase with three moves: bundle decomposition, a per core and core minimum review, and a funded migration alternative held in reserve. Together these reset the conversation from the vendor opening number.
A costed migration plan, even one you never execute, changes who holds the leverage. When Broadcom can see a credible exit with a real budget and timeline, the opening quote moves. The option to walk is worth more than the platform familiarity.
Start 9 to 12 months before the renewal date. That gives time to model the estate, cost an alternative, and run a real internal decision. Buyers who start 60 days out have no leverage and usually pay close to the opening quote.
A credible exit case has a named target platform, a costed project, and a sponsor in the business. It is not a threat in a meeting. It is a board ready plan the vendor can see is real, which is exactly what moves the number.
The reset is not uniform. It scales with core count, with how much of the old stack you actually used, and with whether you were on vSphere Standard or an enterprise agreement. Core dense estates feel it most.
How the transition reset varies by estate profile
| Estate profile | Approx core count | Typical reset band | Main driver |
|---|---|---|---|
| Small | Under 200 cores | 100 to 150% | Loss of vSphere Standard |
| Mid | 200 to 1,000 cores | 120 to 200% | Forced bundle richness |
| Large | 1,000 to 5,000 cores | 150 to 250% | Per core scaling on the full stack |
| Very large | Over 5,000 cores | Negotiated, often lower percent | Volume and exit credibility |
Small and mid estates feel the loss of vSphere Standard most. They are quoted a bundle richer than their workload, so the percentage reset can look as steep as a large estate even though the absolute spend is smaller.
Large estates feel per core scaling on the full stack. With thousands of cores, the VCF subscription multiplies quickly. But these estates also carry the most credible exit threat, which is exactly where negotiated outcomes improve.
Service providers that billed VMware on a usage model face their own reset. The economics that made a hosted VMware service profitable changed, so many are repricing customers or moving workloads. The pressure flows downstream to their tenants.
The Broadcom VMware reset is the steepest single vendor move we tracked in 2026. Most other vendors raised prices in the high single digits to mid twenties. None reset an entire installed base the way the perpetual to subscription conversion did.
Selected 2026 enterprise software price moves, directional bands
| Vendor area | Type of move | Typical effective increase |
|---|---|---|
| Broadcom VMware | Perpetual to subscription reset | 100%+ on perpetual estates |
| Adobe AI tiers | New AI tier and bundle changes | High teens to mid twenties |
| ServiceNow | Renewal uplift and new modules | Low to high teens |
| Microsoft 365 Copilot | AI add on per seat | Roughly doubles a knowledge worker seat |
Other increases sit on top of an existing subscription. The VMware move converts a one time purchase into a perpetual fee. That is a change in the cost model, not a percentage on the same base, which is why it dominates any 2026 price index.
In a multi vendor portfolio, VMware is usually the single largest swing item in 2026. Sequencing matters. Resolve the VMware reset with a clear plan before you spend leverage on smaller renewals elsewhere.
The clauses cap the increase more reliably than the headline discount. A hard annual uplift cap, co terminus dates, swap rights, and a price hold on renewal convert an open ended cost into a bounded one.
The capped annual uplift matters most. It limits how far the subscription can climb at each renewal. Without it, a good first year price can drift upward fast once the initial term ends.
Put the clauses in the master agreement, not a side email. Tie any discount to the cap and the term in one signed document. A discount without a cap is a one year favor, not protection.
Subscription does not remove audit risk. Broadcom can still review deployed cores against entitlements, and a gap becomes a true up bill. Track cores precisely and reconcile before the vendor does, so a routine review never turns into a surprise charge.
Budget the reset as a permanent step up in run rate, not a one time spike. Model three years of subscription cost with the annual uplift, then compare it against a funded migration that starts paying back inside the same window.
Present both paths as run rate over three years, with the break even month marked. Finance leaders respond to a clear crossover point and a funded plan, not to a percentage complaint about the quote.
Sequence the migration by risk and dependency, not by convenience. Move simple, non production workloads first to build the team confidence, then the production estate, leaving the most storage dependent systems for last.
Keep leverage by migrating in visible waves and telling Broadcom the plan is real. Even a partial migration changes the renewal conversation. The credible threat, backed by progress, is what holds the next quote down.
Both paths carry risk. Staying means a higher, recurring cost and growing dependence on a single vendor. Leaving means project cost, operational change, and the chance that the saving is smaller than the disruption. The decision is a balance, not a slogan.
Decide on the three year run rate and the break even point, not on frustration with the quote. If a credible alternative pays back inside two years and your team can absorb the change, the case to move is strong. If not, negotiate hard and stay.
Expect Broadcom to hold the subscription model and lean on the annual uplift rather than another reset. The one time conversion is done for most estates. The next pressure is the renewal, where the uplift and any added modules drive the increase.
The market is responding. Alternatives are maturing, migration tooling is improving, and more buyers are funding exit plans. That competition is the strongest force keeping Broadcom quotes negotiable through 2027.
The reset lands differently by industry because workload density, regulation, and exit options differ. Regulated sectors face the hardest constraints on moving, while lighter estates have more freedom to migrate or renegotiate.
How the reset pressure varies by industry
| Industry | Why the reset bites | Realistic response |
|---|---|---|
| Financial services | Dense estates, strict change control | Negotiate hard, migrate selectively |
| Public sector | Procurement rules, long cycles | Use framework leverage and benchmarks |
| Manufacturing | Edge sites and legacy systems | Right size bundles, phase any move |
| Healthcare | Uptime and compliance constraints | Cap the uplift, migrate low risk first |
Regulated industries carry strict change control and uptime rules, which raise the cost and risk of a migration. That weakens the exit threat, so the negotiated outcome leans more on caps and benchmarks than on a credible move.
Lighter estates with fewer compliance constraints can move faster and use a real exit threat. For them the migration alternative is not just leverage, it is a practical option that often pays back inside the first renewal term.
VMware Cloud Foundation is priced per core, per year, with a minimum of sixteen cores billed per processor. You multiply billable cores across every host by the per core rate, then apply the term and any discount. The minimum is what catches lightly loaded hosts.
Count the physical cores in each processor, apply the sixteen core floor per processor, then sum across every host in scope. Hosts with small processors still bill at sixteen cores each, which inflates the count on older or smaller servers.
Take a host with two processors of twelve cores each. Actual cores total twenty four, but the sixteen core floor bills thirty two. The buyer pays for eight cores they do not physically have, on every host of that shape.
The right questions surface the levers before the contract closes. Ask about the cap, the term, the swap rights, and the support roadmap for any perpetual licenses you keep running.
Pause if the cap lives in an email rather than the contract, if the discount is tied only to the first year, or if swap rights are vague. Each is a sign the protection will not survive the next renewal.
Benchmark the quote against comparable estates by core count, bundle, and term, not against your own last invoice. The reset broke the old reference point, so an external benchmark is the only honest measure of whether a quote is fair.
A benchmark moves the talk from feelings to evidence. When you can show what comparable buyers signed, the opening quote loses its anchor. That is why an independent benchmark is the highest return input to a Broadcom negotiation.
License cost is the visible line, but total cost of ownership includes hardware, power, staff, and support. The reset raises the license line sharply, yet a migration can raise the others, so the honest comparison is full total cost of ownership, not license alone.
A cheaper platform can carry higher staffing and integration costs, especially open source options. The license saving is real, but if it shifts cost onto your team, the total can land flat. Model all the lines before you call it a saving.
Compare three year total cost of ownership for each path, including migration and dual running. The path with the lowest defensible total wins, not the one with the cheapest sticker. This is where many migration business cases quietly fall apart.
The common mistakes are starting late, negotiating the wrong number, and treating the reset as a one off rather than a model change. Each hands leverage to the vendor and widens the gap between list and realized cost.
Avoid them with a calendar, a benchmark, and a funded alternative. None is expensive next to the size of the reset. The discipline is ordinary; the payoff, measured against a six or seven figure increase, is not.
Support and roadmap changes are part of the real cost. After the acquisition, support tiers and account coverage shifted, and some buyers report slower response and fewer named contacts. That operational change has a cost even when it is not on the invoice.
Price the support change as a soft cost in your model and a hard point in negotiation. Ask for defined response times and named coverage in the contract. If the service drops, the value of staying drops with it, which strengthens your exit case.
Governance after the reset means tracking cores, renewals, and usage continuously, not once a year. The subscription model rewards estates that stay right sized and punishes those that drift. A light governance routine protects the saving you negotiated.
Give one owner the mandate for VMware spend across infrastructure, procurement, and finance. Split ownership is how renewals arrive unmanaged and how the uplift compounds. A single accountable owner keeps the estate right sized and the renewal on schedule.
The playbook is a sequence, not a single tactic. Prepare the estate, set a target, build the alternative, then negotiate the clauses in order. Buyers who run the sequence consistently realize the lower end of the increase band.
The strong outcomes share three traits: an early start, a funded alternative, and clauses written into the master agreement. The weak ones share one trait, a late start with no alternative. Preparation, not negotiating flair, drives the result.
A typical transition runs from a surprise quote to a negotiated multi year deal over six to twelve months. The estates that do best treat the quote as an opening position and run a parallel internal process rather than reacting to the clock.
Transitions go wrong when the buyer treats the first quote as final and signs under time pressure. Without an assessment phase, there is no target and no alternative, so the realized number lands close to the opening ask.
Keep the exit option ready by maintaining a current estate model, a named target platform, and a rough order of cost. None requires a migration decision. Together they mean you can produce a credible alternative in days, not months, when the quote lands.
Readiness beats reaction because leverage is about timing. A buyer who can show a credible alternative the week the quote arrives holds the advantage. A buyer who starts modeling after the quote has already lost the most valuable months.
The per core model changes hardware strategy. Denser processors now carry a license cost, so the cheapest server is not always the cheapest to license. Capacity planning and licensing must be decided together, not in separate teams.
When capacity and licensing are planned apart, a refresh that looks efficient on hardware can quietly raise the license bill. Deciding them together lets you size the estate for the lowest combined cost, which is the only number that matters.
The business case compares three year run rate, risk, and effort for each path on one page. It should give leadership a clear recommendation with the break even point, not a pile of data. A crisp case is what unlocks a funded decision.
Get a decision by naming the recommendation and the deadline. Tie it to the renewal date so the choice cannot drift. Leadership decides quickly when the options, the numbers, and the clock are all on a single page.
The standard pitch is that VMware is too embedded to move, so the subscription reset must be absorbed. We disagree. In the Broadcom transitions we have benchmarked, the buyers who funded a parallel migration assessment, even without executing it, held the realized increase far below the opening ask and won multi year caps. The buyer side move is to cost a credible alternative early and let Broadcom see it. The option to walk is worth more than the platform familiarity.
Source: Redress Compliance advisory engagement file, 2024 to 2025.
Broadcom is not selling you a price increase. It is selling you a model change that you can still shape, if you start before the quote arrives.
Related advisory: the Broadcom and VMware advisory practice, the Broadcom VMware licensing pillar, the VMware Cloud Foundation pillar, the Enterprise Software Price Increase Index 2026, and our benchmarking practice.
A perpetual VMware estate converting to subscription typically saw its annual cost reset by more than 100 percent in 2026. That transition reset is a one time repricing. After it, the subscription still carries an annual uplift in the high single digits to low teens at renewal.
No. Broadcom ended new perpetual VMware sales in 2024 and moved the portfolio to subscription only. Existing perpetual licenses keep running but lose new support over time, so the realistic options are a planned migration or a subscription conversion on your terms.
The main bundles are VMware Cloud Foundation, vSphere Foundation, vSphere Enterprise Plus, and a limited vSphere Standard. Broadcom replaced more than 8,000 SKUs with this short list, all sold per core on a subscription. VMware Cloud Foundation is the flagship and the most expensive.
Sometimes. A migration pays off when the multi year license saving clears the one time migration cost and the added operational risk. In our engagements the break even on a credible alternative usually sits 9 to 14 months out, longer for storage heavy estates.
Start with bundle fit, a per core and core minimum review, and a funded migration alternative held in reserve. These three moves reset the conversation from the vendor's opening number. Prepared buyers realized 40 to 60 percent of the opening ask and won multi year caps.
VMware Cloud Foundation is Broadcom's full private cloud stack, bundling vSphere, vSAN, NSX, and Aria, sold per core. Many buyers are quoted it first because it carries the highest price. You are usually not forced into it if a lower tier covers your real workload.
vSphere Standard as a freely available standalone perpetual product was discontinued. It returned in limited subscription form, often only through a partner. Small estates that relied on it are now frequently quoted a richer bundle than their workload needs.
Most production VMware migrations run 9 to 18 months from decision to completion, depending on estate size and storage dependence. Compute heavy estates move faster. The point of a migration plan is not always to execute it, but to hold real leverage at the renewal.
Only if the cap is in the contract. A multi year term with a hard annual uplift cap, co terminus dates, and swap rights is reliable protection. A multi year term with an uncapped renewal at the end simply defers the next increase. The protection is in the clauses.
Most buyers should do both in parallel: negotiate the subscription hard while costing a credible migration. The funded alternative is what caps the subscription. The decision to actually move can wait until the three year numbers are clear and the break even point is known.
The bundle map, the per core reset math, the migration break even model, and the clause checklist that holds the gap widest at renewal.
Built from real Broadcom transitions run since the acquisition closed. Independent. Buyer side. Made for procurement and infrastructure leaders facing the reset.
The estate did not change. The model did. The buyers who saw that early, and costed an exit, paid far less than the opening quote.
