The Broadcom VMware Exit Decision: Ten Moves That Protect the Run Rate

First Broadcom renewal proposals have run 3x to 10x above legacy VMware spend since the acquisition closed in November 2023. These ten moves decide whether you pay it.

Prepared by Redress Compliance  ·  June 2026  ·  Representative VMware estate scenario throughout (benchmark scenario, not a quote)

Every exit decision made without a portability inventory is a guess. Before any platform shortlist, classify every VM by what actually pins it to vSphere: NSX dependence, vSAN dependence, vendor support matrices, latency and compliance constraints, and the licensing of the guest software inside the VM.

The guest software point is the one most teams miss. Oracle and Microsoft server products are licensed per core underneath the hypervisor. Moving those VMs can change their license cost more than the hypervisor swap itself. Run the guest licensing math per target platform before you score the platform.

Tier shares from the representative scenario; they sum to 100 percent of the 8,000 core estate. Benchmark scenario, not a quote.

The inventory is also your negotiation asset. A Broadcom account team can tell the difference between a customer with a slide that says exit and a customer with a tiered inventory and a funded pilot. Only one of them gets a different price.

Select against your tiers, not against analyst quadrants. Five destinations cover nearly every enterprise exit: Nutanix AHV, Microsoft Hyper V, Red Hat OpenShift Virtualization, OpenStack with KVM, and public cloud re platforming. Proxmox earns a mention for edge and branch estates, rarely for the core data center above 200 hosts.

Note what the chart says: against a flat VCF renewal, only two paths beat staying inside three years. The platforms do not win on a three year spreadsheet. They win on years four onward and on what they do to your Broadcom negotiation.

The exit business cases that fail are the ones that hide the parallel run. Migration consumes the savings for two budget years: you pay a bridge subscription on the shrinking VCF estate, the new platform licenses, migration services, retraining, and usually some hardware. Model all of it or the CFO will find it later.

Broadcom publishes no public price list; every number above is an engagement file benchmark at a blended $165 per core per year effective rate. Multi year terms price below one year paper, which matters in section 8. Benchmark ranges: Redress Compliance advisory engagement file, 2024 to 2025.

The decision is a portfolio split, not a platform vote. The inventory tiers from section 1 drive it mechanically.

• Full exit: justified when Tier A and B exceed roughly 85 percent of the estate and no structural NSX or vSAN dependence remains. Highest year one cost, lowest steady state, total Broadcom independence.
• Partial divestiture: the dominant pattern. Migrate Tiers A and B, keep Tiers C and D on a renegotiated residual VCF contract. In the scenario, 5,600 cores move and 2,400 stay. You capture most of the run rate saving and all of the leverage.
• Hold and renegotiate: right when pinned tiers dominate or the business cannot absorb a migration program. Holding is a negotiation strategy only if the exit evaluation is real and documented; otherwise it is a price taker position.

Treat the split as reversible. Wave one teaches you your real migration velocity. Several clients widened the divestiture after wave one beat plan, and one narrowed it when a vendor support matrix refused to certify the target platform.

The transition window is a contract artifact as much as a project plan. Wave order follows the tiers: prove the pattern on Tier A, spend the engineering on Tier B, decide Tier C at the next anniversary, and leave Tier D for the residual contract.

Plan parallel run honestly: budget it per wave, certify applications on the target before cutover, and define rollback criteria in writing. The estates that got into trouble compressed Phase 2 to chase a renewal date they should have negotiated around instead.

The migration and the negotiation are one program with two faces. Broadcom prices conviction. What moved its proposals in our file was never the threat of an exit; it was evidence of one in motion.

Benchmark ranges: Redress Compliance advisory engagement file, 2024 to 2025.

Migrations fail on operations, not on hypervisors. The support plan has five components, and each belongs in a contract, not a slide.

• Parallel run budget: priced per wave in section 3; the bridge subscription line is the cost of safety, not waste.
• Application certification: every Tier B application certified on the target platform before cutover, with the application owner signing the gate.
• Rollback criteria: defined per wave in writing. A wave without a rollback trigger is a bet, not a plan.
• Migration partner terms: fixed fee per wave with acceptance criteria. Time and materials migration contracts inflate exactly when the program is most exposed.
• Skills transition: retraining budgeted in year one ($0.15M in the scenario); the new platform team operational before the last wave, not after it.

The renewal anniversary is the only date Broadcom cares about; make it the date your evidence peaks. Three contract mechanics set the clock.

First, the subscription model has no perpetual fallback. When the term ends, the right to run the software ends with it. An expiry mid migration is unlicensed use, which converts your exit program into audit exposure overnight.

Second, Broadcom applies a surcharge, reported at 20 percent of the first year price, when a renewal lands after the anniversary date. Late is not a negotiating posture with this vendor; it is a fee.

Third, terms are quoted per core with a 16 core minimum per CPU, and Broadcom showed with the announced and then withdrawn 72 core order minimum of April 2025 that packaging rules can change unilaterally between your renewals. Lock sizing rules into the order form, not the price book.

The timing play follows: open the renewal conversation six months before the anniversary with the pilot already funded, size the residual term to your migration schedule rather than to the multi year discount, and never let the anniversary arrive without signed paper on whatever footprint still runs.

An exit program without a quarterly scorecard drifts back to the incumbent. Stand up a steering board across IT, procurement, and finance that tracks four numbers every quarter: cores migrated against plan, VCF cores retired from the bridge contract, run rate delta realized against the section 3 model, and wave schedule adherence.

The scorecard is also negotiation ammunition. A quarterly record of cores leaving the platform is the most credible document you can put in front of a Broadcom account team at the residual renewal. It turns the section 6 evidence ladder into a standing position rather than a one time performance.

The transition period is the highest audit exposure window you will ever run with this vendor. You are operating two platforms, retiring entitlements, and consuming a shrinking subscription. Keep the entitlement record dual platform from day one: subscribed cores, deployed cores, and decommission evidence per cluster, timestamped.

The single most common finding in our file is parallel run usage drifting above subscribed core counts during wave overlap. Reconcile deployed against subscribed cores monthly during Phase 2, and archive host decommission evidence as each wave closes. The Broadcom VMware audit defense briefing covers the full posture.