Broadcom’s aggressive licensing overhaul has set off a chain reaction that could see VMware’s installed base shrink by roughly a third over the next four years, according to Gartner analysts. The research firm’s prediction—that VMware could shed about 35% of its workloads to hyperscalers and alternative platforms by 2028—comes as enterprise IT teams scramble to respond to sweeping changes in how VMware software is sold, renewed, and supported. While the exact figure varies across industry reports, the directional signal is clear: the post-acquisition VMware, now under Broadcom’s tight control, is facing a significant customer exodus, with Microsoft Azure positioning itself as a prime beneficiary.

This article unpacks the forces behind the predicted shift, examines the practical migration pathways available to Windows‑centric organizations, and provides a grounded assessment of the technical, financial, and strategic considerations for IT decision‑makers.

The Broadcom Shock: Licensing and Partner Changes Redraw the Landscape

Broadcom’s acquisition of VMware in late 2023 set the stage for a radical restructuring of the virtualization giant’s commercial model. The changes, many of which took effect in late 2024 and continue to roll out, have upended relationships with customers, resellers, and cloud service providers alike.

Perpetual Licenses Disappear

Perhaps the most jarring move was the immediate halt to the sale of new perpetual VMware licenses. Existing perpetual agreements retain their validity, but all new purchases and renewals now fall under a subscription‑only framework. This shift, while aligning with industry trends, creates a hard break for organizations that historically relied on capital expenditure models and long‑term support relationships. Annual recurring costs are often higher, and the loss of perpetual core entitlements pressures budgets at a time when many IT departments are already stretched thin.

The CSP Program Gutted

The VMware Cloud Service Provider (VCSP) program—the backbone through which thousands of hosting firms, managed service providers, and smaller cloud vendors resold VMware—has been dismantled. By November 1, 2025, existing white‑label and registered reseller pathways will close, replaced by a much smaller, invitation‑only scheme. Broadcom has notified non‑invited partners that their contracts will not be renewed, effectively cutting off a tier of service providers that many mid‑market and SMB customers depend on for VMware‑based cloud and managed offerings. Customers who previously purchased VMware as part of a bundled service from a local CSP must now either buy directly from Broadcom or hope their provider secures an elusive invitation—a gamble that injects uncertainty into renewal cycles.

Reseller Consolidation

On the reseller front, Broadcom has collapsed partner tiers, eliminating “Registered” status and raising qualification thresholds. The result is a smaller cadre of large partners capable of delivering VCF‑centric private clouds, reducing competition and squeezing smaller integrators. For end customers, this consolidation means fewer points of contact, diminished bargaining power, and—by many accounts—higher effective prices.

The 35% Prediction: What Gartner Actually Said

The headline figure—that VMware could lose 35% of its workloads by 2028—originates from a Gartner Symposium presentation and subsequent press reports, notably via SDxCentral. Gartner Vice President Julia Palmer told attendees that the hypervisor addiction “needs to change” and that hyperscalers, Nutanix, and other alternatives were poised to capture a substantial share of the VMware estate. The precise percentage, however, appears in different forms across multiple outlets: some reference 35% of workloads, others speak of up to 30% of the customer base, and equity research firms like William Blair cite market estimates that as much as 30% of VMware’s 400,000+ customers could eventually switch.

These numbers should be treated as directional indicators, not audited metrics. Gartner’s own public-facing analyses stop short of a single, verified dataset; nevertheless, the convergence of estimates around one‑third of the installed base underscores a genuine and widespread customer discontent. The underlying drivers cited by Palmer and industry analysts include:

  • Commercial friction: Higher subscription costs and restricted partner access make alternatives economically compelling.
  • Cloud modernization pull: Hyperscalers offer integrated AI services, serverless computing, and consumption‑based pricing that speed digital transformation.
  • Competitive on‑prem options: Nutanix, Red Hat, and open‑source stacks have matured and are aggressively targeting VMware refugees with simplified licensing and migration tooling.

Where the Workloads Are Heading: Hyperscalers, Nutanix, and Beyond

Microsoft Azure: A Natural Home for Windows Workloads

Microsoft has emerged as perhaps the most aggressive suitor for disaffected VMware customers. Through Azure VMware Solution (AVS), Azure Migrate, Azure Arc, and the Azure Hybrid Benefit, the Redmond giant provides a comprehensive migration stack that preserves familiar vSphere‑based operations while opening a path to cloud‑native modernization.

AVS runs VMware clusters on dedicated, bare‑metal Azure hardware. It is a fully managed service, meaning Microsoft handles patching, lifecycle management, and infrastructure operations. For Windows‑heavy estates, the integration with Azure’s native tooling—Entra ID, Defender for Cloud, Azure Backup—enables a seamless extension of on‑premises identity, security, and governance controls. The Azure Hybrid Benefit allows organizations to bring their existing Windows Server and SQL Server licenses to AVS, significantly reducing compute costs.

However, AVS imposes technical guardrails that require planning. Each private cloud supports between 3 and 16 ESXi hosts per cluster, with a maximum of 12 clusters (96 hosts) per private cloud. For large VMware estates accustomed to hundreds of hosts in a single management domain, this may necessitate splitting workloads across multiple AVS private clouds—a manageable but non‑trivial architectural adjustment. Additionally, AVS is priced at a premium over raw IaaS, so lift‑and‑shift migrations without subsequent optimization can erode the business case over time.

Beyond AVS, Azure Local (formerly Azure Stack HCI) and Windows Server Hyper‑V offer on‑premises hyper‑converged options for shops that wish to move away from VMware but retain complete control of their data center footprint. Hyper‑V, while less favored by Microsoft’s cloud‑first strategy, remains a capable and cost‑effective alternative for smaller deployments and branch offices.

AWS and Google Cloud: Shifting Dynamics

AWS and Google Cloud are also active in the VMware replacement market, though Broadcom’s partner choices have reshaped their roles. Broadcom officially tapped Google Cloud as its first hyperscale partner for VMware Cloud Foundation (VCF) licensing, while AWS—the largest host of VMware Cloud on AWS—was stripped of its license distribution rights. This means customers wanting to run VMware on AWS must now purchase licenses directly from Broadcom and bring them to the platform, a process that undercuts the convenience and pricing simplicity AWS once offered.

Both hyperscalers, however, are courting customers with their own migration accelerators, credits, and managed services. Google’s Google Cloud VMware Engine and AWS’s VMware Cloud on AWS remain viable landing zones, especially for orgs with existing multi‑cloud strategies or specific compliance needs.

Nutanix and On‑Prem Alternatives

Nutanix has emerged as the primary on‑premises beneficiary of VMware turmoil. CEO Rajiv Ramaswami has publicly stated that the company has onboarded “hundreds” of new customers that defected from VMware, and equity research corroborates the trend. Nutanix’s AHV hypervisor and its Acropolis HCI stack promise a familiar, streamlined operational experience with significantly lower licensing costs for many workloads.

Nutanix Move, the company’s migration tool, automates virtual machine conversions from vSphere to AHV, reducing the time and risk of physical‑to‑virtual transitions. But migration complexity should not be underestimated. Large enterprises typically report multi‑year transformation windows, and retraining staff on AHV’s management paradigms—while straightforward—consumes time and budget.

Red Hat, OpenStack, and KubeVirt

For organizations with strong Linux and container expertise, Red Hat’s OpenShift virtualization (built on KubeVirt) and OpenStack present open‑source paths. Gartner’s Palmer noted these options but warned of a skills shortage that could impede adoption. Windows‑centric shops, in particular, may find the learning curve steep and the ecosystem less cohesive than what they’ve come to expect from VMware and Microsoft.

Practical Migration Timeline and Complexity

Any realistic VMware exit is measured in years, not months. Palmer herself cautioned that a full migration can take upward of three years for medium‑to‑large enterprises, factoring in contractual tail periods, application refactoring, and the need to parallel‑run environments. Organizations should expect a 12‑ to 24‑month timeline to move a substantial fraction of their estate, and even longer if they aim to modernize rather than simply lift‑and‑shift.

A Staged Approach

  1. Inventory and classification – Identify all VMs, dependencies, and regulatory constraints. Prioritize workloads that are easiest to re‑host (simple web servers, dev/test environments) or those that gain the most from cloud‑native services.
  2. Pilot migration – Use tools like Azure Migrate, VMware HCX, or Nutanix Move to move a representative set of VMs. Test performance, networking, identity integration, and backup/recovery workflows.
  3. Incremental modernization – For stable workloads, consider replatforming to Azure PaaS services (Azure SQL Managed Instance, App Service) or containerizing with Azure Kubernetes Service once the core lift‑and‑shift is complete. This balances short‑term risk with long‑term agility.
  4. Cutover and rollback planning – Develop detailed runbooks that address DNS cutover, Active Directory synchronization, and storage failback. Network latency and throughput can surprise even experienced teams.
  5. Governance and cost controls – Implement FinOps practices with tools like Microsoft Cost Management and Azure Advisor from day one. Without discipline, hyperscaler bills can overshoot TCO models.

Broadcom’s licensing overhaul has not gone unchallenged. High‑profile lawsuits—including claims from retail and enterprise plaintiffs—allege breach of contract related to perpetual license commitments and support. European trade groups have also petitioned regulators, citing potential anti‑competitive behavior. These legal headwinds inject further uncertainty: a court ruling or regulatory intervention could reshape the timeline or economics for some customers, making it prudent to include contingent clauses in any long‑term migration plan.

What This Means for Windows‑Centric Organizations

Windows‑heavy environments face a unique calculus. Many such estates run decades‑old Windows Server and SQL Server instances tied to VMware vSphere. Moving to Azure unlocks native hybrid licensing benefits that can slash compute costs by up to 40% for Windows Server and up to 55% for SQL Server, providing a compelling financial incentive. Moreover, Microsoft’s deep integration of Active Directory, Group Policy, and System Center with Azure Arc extends on‑premises management models into the cloud without requiring a leap to cloud‑native identity.

However, technical constraints remain. AVS’s cluster size limits may force compartmentalization of large VDI farms or virtualized SQL Server clusters. Latency‑sensitive applications with stringent data locality requirements may be better served by on‑premises Hyper‑V or Azure Local rather than a remotely hosted AVS instance. And for organizations with significant GPU‑accelerated AI workloads, hardware specialization in public cloud may not yet match the density of on‑premises GPU nodes, though Azure’s N‑series VMs are narrowing the gap.

Decision Checklist for IT Leaders

  • Audit contracts and entitlements: Pin down renewal dates, active perpetual licenses, and any obligation to purchase new subscriptions. Model TCO for three scenarios: remain on VMware (VCF), migrate to hyperscaler managed VMware, or replatform to Nutanix/OpenStack.
  • Evaluate migration tooling: Run a proof of concept with Azure Migrate or Nutanix Move on a subset of workloads. Measure true throughput, app compatibility, and operational overhead.
  • Assess staffing and skill gaps: Factor in training for cloud‑native tooling (Terraform, Azure Policy) or new hypervisor interfaces. The soft costs of change management often exceed hardware savings.
  • Negotiate early and often: Hyperscalers and Nutanix are eager to win VMware refugees. Committed migration roadmaps can unlock significant discounts, migration credits, and professional services.
  • Plan for data sovereignty and compliance: For regulated industries, confirm that the target platform supports required certifications (FedRAMP, HIPAA, PCI‑DSS) and data residency guarantees. Azure’s sovereign cloud options (e.g., Azure Government) may be determinative.

Broadcom’s Counter: VCF 9.0 and the Private Cloud Pitch

Broadcom is not standing idle. VMware Cloud Foundation 9.0 introduced earlier this year touts simplified deployment, centralized lifecycle management, and tighter integration with NSX and vSAN. The messaging emphasizes “cost transparency and cost control,” positioning VCF as a competitive private cloud platform that can match hyperscaler economics for certain workloads—particularly those with steady‑state resource demands and sensitive data residency requirements.

Yet, the same commercial moves that angered customers also undercut VCF’s appeal. By concentrating partners and hiking prices, Broadcom has made the total cost of ownership for on‑prem VCF less attractive for all but the largest enterprises with deep VMware expertise. The tug‑of‑war between Broadcom’s VCF ambitions and the centrifugal force of hyperscaler incentives will define the market for the next three years.

Conclusion: Navigating the Shift with Deliberate Speed

The 35% workload displacement figure may be an estimate, but it captures a genuine market sentiment: VMware’s lock on enterprise virtualization is loosening. For Windows‑focused organizations, the path forward is not a binary choice between VMware and chaos; it is a spectrum of options that range from Azure‑managed VMware to full cloud‑native replatforming and on‑premises Hyper‑V.

Success will hinge on deliberate, data‑driven planning. Organizations that inventory workloads, model lifecycle costs, and stage migrations—while wringing concessions from both Broadcom and their chosen destination vendors—will emerge with a more agile, cost‑effective infrastructure. Those that delay risk being forced into reactive decisions at contract renewal time, when leverage evaporates.

The next 18 to 36 months represent a rare moment of high stakes and high opportunity. With clear‑eyed analysis and a pragmatic roadmap, Windows‑centric enterprises can turn the Broadcom disruption into a catalyst for modernization rather than a crisis of control.