Broadcom Is Now Encouraging You to Leave: The 2026 VMware Exit Math After the Renewal Shock

A Nottingham manufacturing firm we spoke to last spring had budgeted around £14,000 for their annual VMware renewal. The quote that came back was £41,000. Same hardware, same workloads, no extra capacity. Their finance director assumed it was a typo. It wasn't.

Since Broadcom completed its acquisition of VMware in late 2023, stories like this have become routine. Perpetual licences — the ones you bought once and kept — are gone, replaced by subscription bundles you may not want and can't unbundle. Support renewals for smaller estates have jumped by anything from 2x to 10x. And Broadcom's leadership has been remarkably open about the strategy: they want the top few thousand enterprise accounts, and they are perfectly happy for everyone else to look elsewhere.

So let's take them at their word. If Broadcom is nudging you towards the door, the sensible response isn't panic or blind loyalty. It's maths. This post lays out a decision framework for working out whether to stay, migrate, or sell the whole evaluation to a client as a standalone piece of work.

First, get the real number in front of you

Before you can decide anything, you need three figures that businesses routinely fail to gather.

The all-in cost of staying. Not just the renewal quote. Add the cost of any bundle features you're now forced to buy but won't use, plus the hardware refresh Broadcom's newer versions may quietly require. If your quote is £41,000 a year, your three-year commitment to staying is roughly £123,000 before you touch a single VM.

The one-off cost of leaving. Migration labour, potential new hardware, staff retraining, and the risk premium of change. This is the number people overestimate through fear and underestimate through optimism, often in the same conversation.

The ongoing cost of the alternative. What does the destination platform cost to run, year after year, once you're there?

Until you have all three, you're arguing about feelings. Once you have them, the decision usually makes itself.

Triage your workloads before you triage your platform

The biggest mistake we see is treating the VMware estate as one thing to be moved wholesale. It almost never is. Sort your virtual machines into four buckets.

Bucket 1 — Retire. Every estate has zombies: VMs nobody has logged into for eighteen months, test environments that outlived their project, the domain controller for a company you acquired in 2019. Killing these costs nothing and shrinks every subsequent calculation. We routinely find 15–30% of a VM count can simply be switched off.

Bucket 2 — Lift-and-shift candidates. Standard Windows and Linux servers doing ordinary jobs: file services, line-of-business apps, print, internal web. These move cleanly to almost any hypervisor. This is where the bulk of your savings live.

Bucket 3 — Cloud-native candidates. Workloads that would genuinely benefit from being reborn as a managed service rather than a VM. A self-hosted mail server is a prime example — moving to Microsoft 365 removes the VM entirely rather than relocating it. Databases that could become Azure SQL, or apps that could run in containers, sit here too.

Bucket 4 — Stay put (for now). Anything with a hardware dependency, a fussy vendor support requirement, or a compliance boundary that makes movement expensive. Some estates keep a small VMware footprint precisely so they don't have to touch these until a natural refresh point.

Only once you've triaged do the platform choices come into focus.

The three realistic destinations

Proxmox VE

For SMEs running on-premises hardware they've already paid for, Proxmox is the most common landing spot. It's open source, based on KVM, and runs happily on your existing servers. There's no per-socket subscription tax. A commercial support subscription for a couple of hosts costs a few hundred pounds a year rather than tens of thousands.

The honest caveats: the management experience is less polished than vCenter, the ecosystem of third-party integrations is thinner, and you're taking on more operational responsibility. For a business with capable internal IT or a managed provider behind it, that trade is easily worth making. For a business with no in-house skills and no partner, it can become a liability.

Azure (or another public cloud)

Moving workloads into Azure changes the shape of the bill from a large capital lump to a monthly operating cost. That suits some businesses and horrifies others. The trap is assuming cloud is automatically cheaper — for steady, always-on servers it frequently isn't. Where cloud wins is in eliminating hardware refresh cycles, absorbing your Bucket 3 workloads into managed services, and giving you elasticity you'd never buy as tin.

Azure VMware Solution exists as a like-for-like refuge, but it carries VMware licensing inside it, so as a cost-escape it's usually a non-starter. The value is in re-platforming, not relocating VMware wholesale.

Hyper-V

Often overlooked. If your estate is Windows-heavy and you already hold the appropriate Windows Server Datacenter licensing, Hyper-V may cost you very little extra and comes from a vendor you already deal with daily. It's not fashionable, but fashion doesn't appear on the invoice.

Your negotiation leverage is real — use it

Even if you intend to stay, don't renew quietly. In the post-perpetual world, Broadcom's account teams still have discretion, and a credible, evidenced threat to leave is your only lever.

That word — evidenced — matters. A vague grumble achieves nothing. A migration plan with costed timelines, a chosen destination platform, and a proof-of-concept already running changes the conversation. Time your renewal discussions early; scrambling three weeks before expiry hands all the power to the vendor. And be willing to walk, because a bluff you're not prepared to execute is worthless the moment they call it.

The overlooked option: sell the evaluation itself

Here's the angle most providers miss. The work of triaging an estate, pricing three destinations, and producing a defensible recommendation has genuine value on its own — separate from whatever migration follows.

For an IT provider, packaging this as a fixed-fee VMware Exit Audit does three useful things. It gives the client a decision they can take to their board with confidence. It gets you paid for the discovery work you'd otherwise give away and hope to recoup later. And it positions you as the trusted advisor for whatever comes next, whether that's a migration you deliver or a negotiation you support.

A typical audit produces a full VM inventory with the four-bucket triage applied, a three-year cost comparison across staying, Proxmox, and Azure, a risk assessment, and a recommended sequence of work. For a business staring at a renewal that's tripled, a few thousand pounds to know the right answer is money well spent — and it often pays for itself in the negotiation alone.

The bottom line

Broadcom has made the decision easier by being blunt: they don't especially want your business. That's oddly freeing. It means loyalty isn't reciprocated and shouldn't factor into your maths.

Sort your workloads. Retire the dead ones. Price the three real destinations against the true cost of staying. Then either move, negotiate from strength, or — if you're the one advising others — sell the clarity itself. The one option that never makes sense is signing the renewal without doing the arithmetic.

If you're facing a 2026 renewal and the number made you wince, that's your cue to start counting. We're happy to help you run those figures before the clock runs down.