Broadcom prices VMware against your exit, not last year's invoice
Broadcom retired VMware perpetual licensing and moved the entire portfolio to subscription, billed per physical core with a minimum core count per processor, inside two bundles. Opening renewal quotes commonly land 2x to 5x the prior perpetual-plus-support cost once the bundle and the minimum are applied, and they settle far lower once an exit plan is credible. The renewal is won before the quote arrives: rationalize the estate, model the right core count, and build a real alternative.
Prepared by Redress Compliance · August 9, 2026 · Broadcom advisory. Based on roughly 30 to 40 VMware renewals benchmarked 2024 to 2025.
Executive summary
The model is per physical core on a term, with a minimum core count per processor, and the minimum is what surprises buyers.
Broadcom counts physical cores per processor and licenses each on a subscription term, so a dense, low-core host can still be charged at the floor: you no longer buy a license you keep, you rent capacity, and sub-minimum hosts pay the 16-core-per-processor floor regardless of real core count.
Adding 10 to 25 percent.
The minimum, not the list rate, is what most buyers miss, and it makes host design a direct factor in the bill, because fewer hosts with higher core counts usually license more efficiently than many small hosts. Model the core count of the planned estate, not the current one.
Most products now sit inside two bundles, so buying a single component is harder and the bundle decision is the largest line on the quote.
VMware Cloud Foundation carries the full stack, compute, storage, network and management, at the highest per-core rate; VMware vSphere Foundation carries the core compute stack at a lower rate.
And many standalone SKUs such as vSAN or NSX were folded in, which pushes buyers toward the larger package whether or not they use it. Buying VCF for an estate that only runs vSphere and vSAN is the most common overspend, with 40 to 70 percent of bundled capability going unused on estates that only needed vSphere and vSAN.
Match the bundle to use, drop VCF to VVF where the workload allows, and keep growth as flexible add-on capacity rather than baked into the base term.
The renewal jump is structural, not a one-off: subscription replaces support, bundle replaces components, minimum replaces actual.
A perpetual estate that paid only annual support now pays a full subscription on every core, inside a bundle, at a core minimum, and each factor compounds the others, so opening quotes commonly land 2x to 5x prior cost, a median 3.1 times in our benchmarks.
That is why the increase cannot be argued away as an error: it is the model working as designed. The three-year prepaid term carries the deepest discount, but it also locks in the bundle before you have rationalized the estate, and a discount on the wrong baseline is still an overspend.
The strongest lever is a credible exit plan, because Broadcom prices against the cost and risk of your migration, not against last year's invoice.
A costed, scheduled migration to an alternative hypervisor or platform changes the conversation because Broadcom knows the alternative is real, and the credibility of the plan, not its execution, is the lever.
Across the benchmarks, buyers cut the opening quote by an average of 38 percent by rightsizing the bundle, consolidating cores to beat the per-processor minimum, and presenting a credible alternative, and the savings came from the baseline, not from a simple volume discount.
Negotiate the bundle and core count before discussing the discount or the term length, and commit multi-year only once the baseline is rationalized.
Broadcom VMware packaging at a glance
| Package | What it includes | Best fit | Watch for |
|---|---|---|---|
| vSphere Foundation (VVF) | Compute virtualization core stack | Smaller estates needing vSphere | Still a per-core minimum |
| Cloud Foundation (VCF) | Full stack: compute, storage, network, management | Large private cloud estates | Paying for unused capability |
| Add-on capacity | Extra cores or features on a bundle | Growth within a term | Mid-term price uplift |
| Multi-year prepaid | Discounted term commit | Stable, rationalized estates | Locking the bundle too early |
The bundle decision is the largest single line on the quote.
VCF carries the full stack and the highest per-core rate; VVF carries the core compute stack at a lower rate; and buying VCF for an estate that only runs vSphere and vSAN is the most common overspend we see, with 40 to 70 percent of bundled capability going unused.
Match the bundle to use, count real cores and consolidate onto fewer higher-core hosts to beat the minimum, and keep add-on capacity flexible rather than baked into the base term.
Because each processor carries a core floor, host design now affects the bill directly: model the core count of the planned estate, not the current one, since fewer hosts with higher core counts usually license more efficiently.
The full negotiation detail sits in the Broadcom VMware negotiation playbook, and why the model changed in why Broadcom killed perpetual licenses.
Why the renewal quote jumped, and the levers that cut it
- Subscription replaces support: a perpetual estate that paid only annual support now rents what it used to own, on every core, which is the base of the increase.
- Bundle replaces components: you pay for the package, not the part, and standalone SKUs folded into VCF and VVF push you toward the larger bundle whether or not you use it all.
- Minimum replaces actual: sub-minimum hosts pay the per-processor floor regardless of real core count, adding 10 to 25 percent, which is why host consolidation is a pricing lever, not just an architecture one.
- Rightsize the bundle and reduce the core count: drop VCF to VVF where the workload allows, and consolidate hosts to beat the per-processor minimum, the two moves that reprice the baseline before any discount.
- Price a credible alternative and hold the term: cost a move to an alternative hypervisor or platform in detail, and commit multi-year only once the baseline is rationalized. The perpetual license options guide and the Hyper-V comparison cover the alternatives.
The Broadcom VMware negotiation playbook
The core subscription math, the bundle pushback, the exit options, and the renewal levers that hold the line on a Broadcom quote.
Get the white paper →How to use a migration plan as leverage
Broadcom does not price VMware against last year's invoice; it prices against the cost and the risk of you leaving, so your strongest number is your exit plan.
An exit plan does not have to be executed to be useful: a costed, scheduled migration changes the conversation because Broadcom knows the alternative is real, and the credibility of the plan, not its execution, is the lever. The renewal is therefore won before the quote arrives, in three moves.
Rationalize the estate, mapping which workloads genuinely need the full VCF stack and which only need VVF compute, because 40 to 70 percent of bundled capability goes unused on estates that only ran vSphere and vSAN.
Model the right core count, consolidating onto fewer higher-core hosts so the per-processor minimum stops charging for cores you do not physically run, and modelling the planned estate rather than the current one.
And build a credible alternative, costing a move to an alternative hypervisor or platform in detail including effort and risk, because that is the number Broadcom actually prices against.
Then request the renewal quote broken down by bundle, core and term line by line, and negotiate the bundle and core count before discussing the discount or the term length, committing to a multi-year term only once the baseline is rationalized.
The common advice to sign a three-year prepaid quickly to lock the deepest discount is exactly backwards, because a deep discount applied to an unrationalized estate locks in unused VCF capability and an inflated core count for three years, and a discount on the wrong baseline is still an overspend.
The pricing-report detail sits in the Broadcom VMware pricing report.
- Percentile standing for your exact deal size and industry, from real closed transactions
- Scenario simulation before the call: test alternative terms and see the financial impact of each
- A negotiation playbook, talking points, and a two page executive brief on day one
What we saw across Broadcom VMware engagements, 2024 to 2025
Across roughly 30 to 40 VMware renewals Morten Andersen benchmarked between 2024 and 2025, the first quote almost always priced the whole estate as one bundle at full core count, and the common advice made it worse.
The standard advice is to sign a three-year prepaid term quickly to lock the deepest discount before prices rise again. We disagree:
The median opening renewal quote against prior perpetual-plus-support cost, once the bundle and the per-core minimum were applied to the whole estate.
Average reduction achieved by rightsizing the bundle, consolidating cores, and presenting a credible migration alternative, from the baseline not a volume discount.
In roughly half the renewals, the deep multi-year discount was applied to an estate that had not been rationalized, so buyers locked in unused VCF capability and an inflated core count for three years, and a discount on the wrong baseline is still an overspend.
Three patterns recurred: bundle inflation, with 40 to 70 percent of bundled capability unused on estates that only needed vSphere and vSAN; core minimum drag, with dense hosts billed for a 16-core-per-processor minimum they did not physically reach, adding 10 to 25 percent.
And anchored quotes, with opening figures at 2x to 5x prior cost that settled far lower once an exit plan was credible.
The buyer-side sequence is to inventory every host, processor and physical core, map which workloads need VCF and which only need VVF, model a consolidated host design that minimizes the per-processor minimum, cost a credible migration to an alternative platform.
Request the quote broken down line by line, negotiate the bundle and core count before the discount or the term, and commit multi-year only once the baseline is rationalized.
A credible migration alternative moves a Broadcom quote more than any volume argument, because it reprices the renewal against your exit cost. The wider library sits in the VMware practice.
Your first five moves
- Inventory every host, processor and physical core across the estate, because the per-core minimum and the bundle both price against a count you have to verify before Broadcom rounds it up.
- Map which workloads need VCF and which only need VVF, and size the bundle to use, because buying VCF for a VVF workload is the most common overspend.
- Model a consolidated host design that minimizes the per-processor minimum, because fewer higher-core hosts license more efficiently than many small ones.
- Cost a credible migration to an alternative platform, including effort and risk, because that exit number is what Broadcom actually prices against.
- Negotiate the bundle and core count before the discount or the term, and commit multi-year only once the baseline is rationalized. The VMware practice runs the renewal with you.
Frequently asked questions
How is VMware priced after the Broadcom acquisition?
VMware is now subscription only, priced per physical processor core, with a minimum core count applied to each processor.
Perpetual licensing has ended, so you rent capacity on a one, three, or five year term rather than buying a license you keep, with the deepest discount on prepaid multi-year commitments.
Most products sit inside the VCF and VVF bundles, so buying a single component outside a bundle is now harder, which pushes buyers toward the larger package.
What is the VMware per-core minimum?
Broadcom applies a minimum core count to each processor, commonly 16 cores, so a dense host with fewer physical cores is still charged at the floor. The minimum, rather than the list rate, is what most surprises buyers, and it adds 10 to 25 percent on sub-minimum hosts.
It makes host design a direct factor in the bill: fewer hosts with higher core counts usually license more efficiently, so model the planned estate's core count, not the current one.
What is the difference between VCF and VVF?
VMware Cloud Foundation bundles the full stack of compute, storage, networking and management at the highest per-core rate, for large private-cloud estates. VMware vSphere Foundation bundles the core compute stack at a lower rate, for smaller estates needing vSphere.
Buying VCF for a workload that only needs VVF is the most common overspend, with 40 to 70 percent of bundled capability unused on estates that only ran vSphere and vSAN, so match the bundle to actual use.
Why did my VMware renewal quote increase so much?
The increase is structural, not a one-off.
A perpetual estate that paid only annual support now pays a full subscription on every core, inside a bundle, at a core minimum, and each factor compounds the others, so opening quotes commonly land two to five times the prior cost, a median 3.1 times in our benchmarks.
It cannot be argued away as an error because it is the model working as designed, which is why the lever is the baseline, the bundle and the core count, not a complaint about the number.
How much can you save on a Broadcom VMware renewal?
In our 2024 to 2025 benchmarks, buyers cut the opening quote by an average of around 38 percent by rightsizing the bundle from VCF to VVF where the workload allowed, consolidating cores to beat the per-processor minimum, and presenting a credible migration alternative.
The savings came from the baseline, not from a simple volume discount, so the work is done before the quote arrives, and a credible exit plan moves the number more than any volume argument.
What is the strongest negotiation lever with Broadcom?
A credible, costed exit plan. Broadcom prices the renewal against the cost and risk of your migration, not against last year's invoice, so a costed, scheduled migration to an alternative hypervisor or platform reprices the conversation.
The plan does not have to be executed to be useful, but it does have to be real, because the credibility of the alternative, not its execution, is the lever that moves the quote.
