VMware Cloud Foundation 5.2 lists at $350 per core per year in 2026, with VVF (vSphere Foundation) at $135 per core, and Broadcom's minimum 16-core-per-CPU commit raising the effective starting point to $5,600 per socket per year for VCF and $2,160 for VVF. A 1,500-core enterprise estate that paid $400,000 per year on perpetual VMware support pre-Broadcom now lists at $525,000 per year on VCF subscription. Realised post-discount pricing lands between $185 and $275 per core for VCF and $80 to $110 for VVF at enterprise scale. This pillar covers the post-Broadcom VCF and VVF tier structure, the minimum-core commit math, the bundle inclusion shifts, and the negotiation levers that move VCF deals in 2026.
Inside This Pillar
- The post-Broadcom VCF portfolio
- Per-core pricing mechanics
- The 16-core minimum commit
- VCF versus VVF decision
- NSX and vSAN bundling
- Perpetual to subscription conversion
- Discount bands by deal size
- Alternative platform pressure
- The vCPU versus per-core debate
- Negotiation levers for VCF
- VCF exit strategy
- How to reduce VCF cost
The post-Broadcom VCF portfolio
Broadcom consolidated more than 8,000 historic VMware SKUs into two flagship subscription products in 2024: VMware Cloud Foundation (VCF) and VMware vSphere Foundation (VVF). A small number of standalone SKUs remain for vSphere Standard, vSphere Essentials Plus Kit, and partner ecosystem cases.
| Product | List per core per year | What it includes |
|---|---|---|
| VMware Cloud Foundation (VCF) | $350 | vSphere, vSAN 1TiB per core, NSX, Aria Suite, HCX, Tanzu Kubernetes |
| VMware vSphere Foundation (VVF) | $135 | vSphere, vSAN 100GiB per core, Aria Standard, Tanzu Kubernetes Standard |
| vSphere Standard | $50 | Compute virtualisation only, no storage/networking inclusion |
| vSphere Essentials Plus Kit | ~$650 per CPU one-time | SMB bundle, 3 hosts, 2 CPUs each |
| VMware Live Recovery | $50 per core add-on | Disaster recovery and ransomware recovery |
The collapse from 8,000 SKUs to two flagship products simplifies procurement and forces customers into an all-or-nothing bundling decision. Customers who previously bought vSphere Enterprise Plus standalone, or NSX standalone, now pay for the bundled inclusion whether or not they use the included products.
Per-core pricing mechanics
Broadcom shifted VMware pricing from per-CPU to per-core in 2024. The per-core model removes the artificial CPU bracketing that previously incentivised dense-core CPU selection. The price scales linearly with the physical core count of the hosts in the cluster.
A 32-core host on VCF lists at $11,200 per year ($350 per core). A 64-core host doubles to $22,400. A 96-core host triples to $33,600. The economics of dense-core CPU selection that drove customers toward 96-core AMD EPYC or 64-core Intel Xeon Platinum chips now run in the opposite direction. Lower-core-count CPUs running at higher utilisation often produce better dollars-per-workload outcomes.
The 16-core minimum commit
The most disruptive Broadcom pricing change is the 16-core minimum commit per CPU socket. A customer running 8-core CPUs in a host still pays for 16 cores per CPU even if only 8 physical cores exist. The effect is most pronounced on older hardware refreshing into VCF subscriptions.
| Physical cores per CPU | Billable cores per CPU | Overcommit impact |
|---|---|---|
| 8 | 16 | 100 percent |
| 10 | 16 | 60 percent |
| 12 | 16 | 33 percent |
| 16 | 16 | 0 percent (no overcommit) |
| 24 | 24 | 0 percent |
| 32 plus | Actual | 0 percent |
Customers with older hardware face a refresh decision earlier than they expected: either replace 8-core or 10-core CPUs with 16-core or denser CPUs to align with the commit floor, or absorb the overcommit penalty until the next hardware refresh cycle. The refresh decision interacts with the alternative platform decision (Nutanix, Red Hat OpenShift, Microsoft Azure Local) detailed below.
VCF versus VVF decision
VCF at $350 per core lists at 2.6 times VVF at $135. The decision between them is the bundle inclusion delta. VCF adds: full NSX networking (versus none in VVF), Aria Suite for cloud operations (versus Aria Standard in VVF), HCX for migration, full Tanzu Kubernetes Grid (versus Standard tier in VVF), and 1 TiB of vSAN per core (versus 100 GiB in VVF).
For customers using NSX in production, VCF is typically the correct choice because standalone NSX is no longer sold. For customers not using NSX or considering its replacement with cloud-native networking, VVF is meaningfully cheaper.
The honest scrutiny: most VCF customers use less than 30 percent of the bundle inclusion. NSX adoption is partial. Aria Suite is partial. Tanzu Kubernetes adoption is partial. Customers paying for VCF for vSphere alone overpay by $215 per core per year.
NSX and vSAN bundling
NSX, vSAN, and Aria Suite are not separately sellable in 2026. They live inside VCF or VVF. Customers who previously bought NSX standalone for $1,200 per CPU per year now pay for it inside the VCF $350 per core whether or not they deploy NSX.
vSAN inclusion differs by tier. VCF includes 1 TiB of vSAN per licensed core. VVF includes 100 GiB. Customers requiring more vSAN capacity pay per-TiB add-on at approximately $20 to $35 per TiB per month depending on storage class.
The negotiating lever on bundling: Broadcom account teams maintain limited but real flexibility on the NSX and vSAN attach for customers who explicitly do not use the included capability. Document non-use at contract negotiation.
Perpetual to subscription conversion
Customers transitioning from VMware perpetual licenses with support contracts to VCF subscription face a step-change cost increase. The pattern across the advisor cohort:
| Prior position | Subscription cost increase |
|---|---|
| vSphere Enterprise Plus on support | 2.5x to 4.5x |
| Full VCF perpetual on support | 1.4x to 2.2x |
| vSphere Standard on support | 3x to 6x (forced bundle upgrade) |
| NSX Enterprise Plus standalone | Variable, often 0.8x to 1.4x (gets vSphere free in the bundle) |
The largest step-change increases hit customers who ran vSphere Standard with point products bolted on, because the simplified portfolio forces them into VCF or VVF whether or not they need the bundle. The smallest increases hit customers who already paid for a near-VCF perpetual stack.
The Broadcom conversion negotiation pattern: Broadcom account teams routinely accept conversion-period discounts of 25 to 45 percent off list to bridge customers from perpetual to subscription. The discount is conditional on multi-year term, full estate commit, and signed conversion within a fixed window. Customers who wait beyond the conversion window pay closer to list. Customers who commit at the window peak with documented competitive alternatives extract the deepest pricing.
Discount bands by deal size
Advisor-observed VCF discount bands for 2026 negotiations.
| Core count | 1-year | 3-year | 5-year |
|---|---|---|---|
| 500 to 2,000 cores | 15 to 25 percent | 22 to 32 percent | 28 to 38 percent |
| 2,000 to 10,000 cores | 22 to 32 percent | 30 to 42 percent | 36 to 48 percent |
| 10,000 to 50,000 cores | 30 to 42 percent | 38 to 52 percent | 44 to 58 percent |
| 50,000 plus cores | 38 to 52 percent | 45 to 60 percent | 50 to 65 percent |
The discount distribution within each band depends on competitive alternative documentation, refresh cycle alignment, and existing perpetual position.
Alternative platform pressure
Five credible alternative platforms move VCF pricing in 2026: Nutanix AHV, Microsoft Azure Local (formerly Azure Stack HCI), Red Hat OpenShift Virtualization, Proxmox Virtual Environment, and the hyperscaler-native options (AWS, Azure, GCP migration with workload refactoring). See VMware alternatives for the deep evaluation framework. Each alternative carries different switching cost, capability gap, and operational reality, so the analysis is genuinely customer-specific.
The vCPU versus per-core debate
Customers running mixed core utilisation (workloads with low average vCPU consumption per host) have argued for a vCPU-allocation pricing model. Broadcom has rejected this for now. The result is that customers running 200 vCPUs across a 64-core host pay for the 64 physical cores rather than the lower vCPU consumption.
The optimisation response: consolidate workloads onto fewer hosts to align physical core count with workload need. Run higher utilisation per host. This reduces the licensable core count even though it does not change the workload footprint.
Negotiation levers for VCF
Seven levers move VCF and VVF commercial terms in 2026.
Multi-year commit. Three-year commits deliver 8 to 12 additional discount points. Five-year commits add 4 to 8 more. Cap the annual escalator at 3 to 5 percent.
VVF instead of VCF. Where NSX and the full bundle are not deployed, VVF is the correct purchase. Documented non-use of NSX, Aria, and Tanzu opens the VVF tier without losing operational capability.
Workload consolidation pre-renewal. Reduce billable core count by consolidating workloads onto denser hosts. Often produces 15 to 25 percent core reduction.
Competitive alternative documented. Nutanix, Azure Local, or OpenShift Virtualization documented evaluation moves VCF pricing materially.
Conversion-window timing. Sign conversion at the peak of Broadcom's conversion-credit window rather than after it closes.
Hardware refresh alignment. Refresh CPUs to 16-core or denser ahead of conversion to eliminate the minimum-commit overcommit penalty.
Estate split. Consider whether a portion of the estate should move to a non-VMware platform, leaving only the workloads where VCF is genuinely the best fit. This dual-platform strategy reduces commit volume on Broadcom.
VCF exit strategy
Full VCF exit is feasible but costly. The typical 1,500-core mid-market VMware estate carries a $1.8M to $5M exit cost across replatform engineering, hardware refresh, and operational rebuild. The exit pays back in 18 to 36 months at enterprise discount levels on the alternative. Below 24 months payback, exit is straightforward. Above 36 months payback, partial exit (move only the workloads with cleanest fit) is the more defensible economic answer.
See VMware alternatives and VMware Broadcom guide for the deeper exit analysis.
How to reduce VCF cost
Pre-renewal (12 to 18 months). Audit actual NSX, vSAN, Aria, and Tanzu utilisation. Identify VVF candidates. Consolidate workloads onto denser hosts to reduce billable core count. Refresh CPUs to 16-core minimum. Run a documented Nutanix or Azure Local evaluation.
At renewal. Apply the seven levers above. Negotiate the multi-year term with escalator cap. Bundle the conversion-window credits where they remain available. Document the competitive alternative as a written internal decision.
Mid-term. Continue consolidation discipline. Track workload migration to hyperscaler-native options for the workloads where that is the right destination, reducing VCF estate size at the next renewal.
For the broader VMware framework, see VMware Broadcom guide, VMware alternatives, Broadcom changes, Broadcom support reduction strategies, VMware licensing consulting buyers guide, negotiating Broadcom, and VMware migration. The VMware and Broadcom vendor hub aggregates the cluster. Engagement starts at cloud contract negotiation or software licensing advisory.