What is VMware Cloud Foundation (VCF)? Definition, Components, and the 2026 Alternatives

VMware Cloud Foundation is a private cloud platform that integrates VMware's virtualization, storage, networking, and Kubernetes capabilities into a single subscription product. VCF is positioned by Broadcom as the strategic direction for VMware infrastructure customers, replacing the historical model of separately licensed vSphere, NSX, vSAN, and other VMware products.

At its core, VCF bundles VMware's foundational infrastructure components: vSphere ESXi as the hypervisor, NSX for software-defined networking and security, vSAN for hyperconverged distributed storage, Avi Load Balancer (formerly NSX Advanced Load Balancer) for application delivery, vDefend for distributed firewall and IDS/IPS, and VMware vSphere Kubernetes Service (VKS) for Kubernetes integration. The components are designed to work together as a unified stack rather than as separately purchased products that customers integrate.

The architectural model behind VCF is the software-defined data center (SDDC). The idea is that compute, storage, networking, and security can all be virtualized in software running on standard x86 servers, with the SDDC Manager component orchestrating the deployment, lifecycle, and operations of the integrated stack.

VCF has been positioned by VMware and Broadcom in various ways over its history. Originally launched as a turn-key SDDC deployment platform, it has evolved into the primary commercial vehicle for VMware infrastructure. As of 2024, organizations purchasing VMware infrastructure typically purchase VCF rather than the individual components. Standalone vSphere licensing remains available but is increasingly less common in new commercial discussions.

The practical definition for a practitioner is that VCF is the modern VMware infrastructure bundle. When someone refers to a VMware-based private cloud or to a VMware deployment with software-defined networking and HCI storage, they are most likely describing a VCF deployment. The underlying technologies are familiar (ESXi, NSX, vSAN), but the commercial vehicle and the integration model is VCF.

Understanding VCF in 2024 requires understanding the commercial context that has dominated the conversation since late 2023.

Broadcom completed its acquisition of VMware in November 2023. The acquisition was strategically positioned by Broadcom as a refocusing of VMware on its highest-value customers and on its core infrastructure platform, with the expectation that revenue from those customers would justify the acquisition price.

The post-acquisition changes were substantial. Perpetual licenses for VMware products were eliminated in favor of subscription-only commercial terms. The product portfolio was consolidated, with many separately licensed products being bundled into VCF as the primary commercial vehicle. Channel partner relationships were restructured, with smaller partners losing direct VMware status. Customer-specific pricing was renegotiated at renewal, often with significant increases.

The customer experience of these changes has varied. Some customers, particularly large enterprises with substantial VMware footprints and existing strategic partnerships with VMware, have experienced changes that they consider acceptable. Other customers, particularly mid-size organizations and those whose specific product mix did not map cleanly to the new VCF bundles, have experienced renewal pricing that is multiples of their prior costs.

The market response has been a measurable shift toward evaluating alternatives. Vendors offering alternative virtualization platforms (Nutanix, Microsoft, others) have reported substantial increases in evaluation activity from VMware customers. Industry analysts have published research documenting the shift. The "VMware exit" has become a recognizable category of project that many organizations are actively executing or evaluating.

VCF bundles several VMware products into a single subscription. Understanding what each component does clarifies what the customer is purchasing.

The hypervisor that runs virtual machines. ESXi is the foundational component on which the rest of VCF is built. It provides VM lifecycle management, resource scheduling, live migration through vMotion, and high availability. ESXi is the longest-tenured component in the VMware portfolio, dating back to the early 2000s, and is mature and widely deployed.

The management plane for ESXi clusters. vCenter handles VM lifecycle management, host configuration, storage policy management, network configuration for Distributed Virtual Switches, and integration with other VMware products. vCenter is included with VCF and is the operational interface that administrators use most frequently.

Software-defined networking for VMware environments. NSX provides overlay networks (Geneve encapsulation), distributed routing, distributed firewalling, microsegmentation policies, and various advanced networking capabilities. NSX has been VMware's network virtualization product since the acquisition of Nicira in 2012 and has gone through multiple major architecture revisions.

Distributed storage for VMware HCI deployments. vSAN runs in the ESXi kernel and aggregates local storage on each host into a cluster-wide distributed storage pool. vSAN supports deduplication, compression, erasure coding, and various data services. vSAN ESA (Express Storage Architecture) is the current generation that uses NVMe-optimized data layouts for higher performance.

Avi Load Balancer (formerly NSX Advanced Load Balancer) provides software-defined load balancing for application traffic as a software function rather than requiring dedicated hardware appliances. vDefend provides distributed firewall and intrusion detection/prevention for east-west security controls within the data center. VMware vSphere Kubernetes Service (VKS) allows Kubernetes clusters to be deployed and managed as part of the VCF environment. SDDC Manager orchestrates deployment, configuration, patching, and upgrade across the VCF components.

VCF 5.2 was released in late 2024, positioned as a release focused on production AI workloads, Kubernetes scale, security, and operational efficiency. Understanding what is new in 5.2 clarifies Broadcom's current strategic positioning for the platform.

The AI workload positioning is central to the 5.2 release. Broadcom is positioning VCF 5.2 as a unified platform for production AI inference, agentic AI applications, and traditional workloads on the same infrastructure. The argument is that AI workloads benefit from running on a platform that provides VM, container, and AI service capabilities through a single management plane rather than requiring separate AI-specific infrastructure.

Specific AI-related capabilities in VCF 5.2 include support for NVIDIA ConnectX-7 NICs and NVIDIA BlueField-3 with Enhanced DirectPath I/O for high-speed multi-host AI training and inference, multi-accelerator GPU choice across AMD and NVIDIA, support for AMD and Intel CPU platforms in mixed compute configurations, NVMe memory tiering for workload density on existing infrastructure, and integration with the broader VMware Private AI capabilities.

Kubernetes capabilities expanded significantly. VKS in 5.2 includes improved cluster scale, faster deployments, and shorter upgrade windows compared to previous versions. Simplified Container-as-a-Service makes VKS more accessible for application teams. The Tanzu Marketplace integration provides certified middleware paths.

Security capabilities focus on zero-trust positioning. vDefend extends distributed IDS/IPS protection to Kubernetes AI workloads. The release delivers enhanced threat inspection performance for distributed inference and improved application identification. Zero-downtime live patching supports many use cases without host evacuation. Continuous compliance enforcement provides automated audit readiness.

VCF licensing is subscription-based, per-core, with editions that provide different feature sets. The specific commercial terms are negotiated with Broadcom or with Broadcom's authorized resellers.

The subscription model means customers pay annually (or on multi-year subscription terms) for the right to use VCF. Perpetual licenses, which were available in earlier VMware commercial history, are no longer offered. Subscription-only is the only available model under Broadcom.

The per-core pricing has minimum core counts per CPU and minimum core counts per cluster in some configurations. The specific minimums vary by edition and have changed across Broadcom's commercial revisions. Organizations evaluating VCF should understand the minimums for their specific configuration because they can drive the licensing cost above what a simple per-core calculation might suggest.

VCF is offered in editions with different feature sets. Specific edition names have evolved through Broadcom's commercial revisions. The general pattern is that higher-tier editions include more components and capabilities. Lower-tier editions provide foundational vSphere with some bundling.

The commercial reality for most organizations evaluating VCF in 2024 is that the renewal price is the relevant comparison. The list price for VCF is one data point, but the actual price the customer will pay depends on the negotiated subscription terms. Organizations should engage Broadcom or authorized partners early in their evaluation to understand their specific renewal economics.

VCF is well-suited to specific environments and less suited to others. Identifying the fit profile clarifies whether VCF should be the strategic platform or whether alternatives deserve serious evaluation.

VCF fits well in organizations that have deep existing VMware investment that is strategically aligned. Large enterprises with mature vSphere environments, extensive NSX-T deployments, mature SRM-based DR, and PowerCLI-based automation have substantial investment in the VMware ecosystem. For these organizations, the migration cost to alternatives is high, and the operational continuity benefits of staying on VMware are substantial. VCF is the strategic platform direction within VMware for these organizations.

VCF fits in environments that use multiple VMware products together. Organizations using vSphere plus NSX plus vSAN plus vRealize plus SRM are exactly the customer profile that the VCF bundle was designed for. The integrated platform model has operational benefits for these multi-product deployments that may justify the subscription cost.

VCF fits in AI workload environments where the Broadcom positioning resonates. Organizations actively deploying production AI inference, agentic AI applications, and mixed CPU and GPU workloads who have evaluated VCF 5.2's AI capabilities and find them compelling can position VCF as the AI platform.

VCF fits less well in environments that have minimal existing VMware investment. Greenfield deployments without legacy VMware footprint have no obligation to choose VMware. The case for VCF in greenfield is harder to make against alternatives that may have lower commercial cost and simpler operational models.

VCF fits less well in environments where the licensing economics have become operationally difficult. Organizations whose renewal pricing has increased significantly may find that the cost-benefit calculation no longer supports remaining on VMware. These organizations are the typical VMware exit candidates.

VCF fits less well in environments with a narrow VMware product footprint. Organizations using only vSphere ESXi without NSX, vSAN, or other bundled products receive the bundle but only use a small portion of it. The economic justification for the bundle is weaker when the customer is paying for capabilities they do not use.

For organizations evaluating alternatives to VCF, the realistic options in 2024 are limited to a small number of platforms. Understanding what each offers is the foundation for the evaluation.

Nutanix Cloud Infrastructure (NCI) with AHV is the most common destination for organizations exiting VMware. Nutanix has been positioned commercially as the VMware exit alternative and has captured the substantial majority of announced VMware migrations. NCI provides hyperconverged infrastructure with AHV as the included hypervisor (no separate hypervisor licensing), distributed storage through DSF, and integrated management through Prism. The Nutanix platform supports both full HCI deployments and disaggregated deployments where AHV runs on compute-only nodes connected to external Pure FlashArray storage.

Microsoft Azure Stack HCI with Hyper-V is the alternative for organizations with strong Microsoft and Azure alignment. Azure Stack HCI is Microsoft's hyperconverged infrastructure platform running Hyper-V as the hypervisor and Storage Spaces Direct as the distributed storage. The platform integrates with Azure for hybrid cloud scenarios, Azure Arc for management, and various Azure services for backup and DR. Azure Stack HCI is most commonly chosen by organizations with existing Microsoft enterprise agreements and operational alignment with Microsoft.

Amazon VMware Cloud on AWS is a bridge approach rather than a full alternative. VMware Cloud on AWS runs VMware vSphere in AWS, allowing organizations to decommission on-premises VMware hardware while continuing to run VMware workloads in AWS during a transition period. VMware Cloud on AWS is BYOL (Bring Your Own License) meaning the customer continues to pay Broadcom for the VMware licensing while paying AWS for the infrastructure. VMware Cloud on AWS does not eliminate VMware licensing cost; it changes the underlying infrastructure from on-premises to AWS.

Other platforms exist but are less commonly chosen for VMware exits. OpenStack with KVM is technically capable but operationally complex and has limited commercial support options for enterprise customers. Proxmox VE has a growing footprint particularly in cost-sensitive segments but has limited enterprise track record. Various smaller HCI platforms exist but have either limited enterprise scale or limited VMware exit migration tooling.

Nutanix Cloud Infrastructure with AHV is the most commonly chosen alternative to VCF for organizations executing VMware exits. The architectural fit is strong because both platforms address the same use cases through different implementations.

NCI provides the foundational platform: AHV as the hypervisor, the Distributed Storage Fabric as the storage layer, and Prism as the management plane. AHV is included in the NCI license at no additional hypervisor cost, which is the structural economic advantage compared to VCF where ESXi licensing is part of the VCF bundle.

For HCI deployments, NCI competes directly with VCF on hyperconverged infrastructure. The architectural model is similar (software-defined storage running on standard servers, integrated with the hypervisor) but the implementations differ. Nutanix uses CVMs for the storage layer; vSAN runs in the ESXi kernel. The operational behavior is broadly similar for most workloads.

For disaggregated deployments, NCI with AHV running on compute-only nodes with external Pure FlashArray storage (the FlashStack with Nutanix pattern) provides a path that VCF does not directly address. VCF's primary architecture is HCI through vSAN. The disaggregated architecture is one of the more meaningful architectural differentiators between Nutanix and VMware in 2024.

For organizations with significant NSX-T investment, the migration to Nutanix is more complex. AHV does not support NSX-T. The equivalent capability on Nutanix is Flow Networking, which has a different policy model and configuration approach. Migration of NSX-T configurations to Flow Networking is non-trivial and is the primary reason that some VMware exits stage their migration with Amazon VMware Cloud on AWS as a bridge for NSX-dependent workloads.

The licensing comparison favors Nutanix in most evaluations. The specific dollar comparison depends on the customer's VMware renewal terms and Nutanix's specific pricing, but the structural advantage of AHV being included in the NCI license (rather than ESXi being a separate licensed component within VCF) typically produces favorable economics.

Microsoft Azure Stack HCI is the alternative platform for organizations with strong Microsoft alignment. The fit is specific and is less common than Nutanix for VMware exits but is the right choice for specific environments.

Azure Stack HCI runs Hyper-V as the hypervisor with Storage Spaces Direct as the distributed storage layer. Both are mature Microsoft technologies with substantial production deployment. The platform is positioned by Microsoft as the on-premises companion to Azure, with deep integration to Azure services.

The architectural model is hyperconverged. Storage is distributed across the cluster nodes, with each node contributing both compute and storage capacity. The model is analogous to VMware vSAN within VCF and to Nutanix HCI.

The integration with Azure is the primary differentiator. Azure Arc enables management of Azure Stack HCI clusters as if they were Azure resources, with consistent management surfaces, monitoring, governance, and deployment workflows. Azure backup, Azure Site Recovery, and other Azure services have native integration with Azure Stack HCI.

For organizations with existing Microsoft enterprise agreements, the licensing economics can be favorable. Hyper-V is included in Windows Server, which many enterprises already license. Azure Stack HCI adds a per-core subscription on top of Windows Server licensing. The combined cost depends on the customer's specific Microsoft licensing structure.

For organizations with strong operational alignment to Microsoft (Active Directory, System Center, Microsoft 365, Azure as the primary cloud), the platform fits naturally. The operational tooling is familiar. The vendor relationship is established. The integration with Microsoft enterprise platforms is direct.

For organizations without strong Microsoft alignment, Azure Stack HCI is less commonly chosen. The platform's strengths are most meaningful in Microsoft-aligned environments. In environments where Microsoft is not the primary platform vendor, Azure Stack HCI does not provide differentiated value.

Amazon VMware Cloud on AWS is a specific architectural pattern that organizations use as part of a VMware exit strategy rather than as a destination platform. Understanding what VMware Cloud on AWS is, and what it is not, clarifies its role.

VMware Cloud on AWS runs VMware vSphere in AWS infrastructure. From the workload perspective, VMware Cloud on AWS looks like vSphere. Workloads run on ESXi. vCenter manages the environment. NSX provides network virtualization. The familiar VMware operational model is preserved.

What VMware Cloud on AWS is not: VMware Cloud on AWS is not Amazon EC2 instances. VMware Cloud on AWS is a specific VMware-and-Amazon partnership that delivers vSphere running on AWS-managed infrastructure as a service.

VMware Cloud on AWS is BYOL. The customer provides Broadcom licensing for the vSphere environment running in VMware Cloud on AWS. The customer also pays AWS for the underlying infrastructure capacity. The combined cost is the AWS infrastructure cost plus the Broadcom licensing cost.

The economic implication is important. VMware Cloud on AWS does not eliminate Broadcom licensing. Customers still pay Broadcom. VMware Cloud on AWS does not reduce VMware costs. Customers pay for both Broadcom and AWS. The use case for VMware Cloud on AWS is therefore not cost reduction.

The use case for VMware Cloud on AWS is bridging. Organizations that need to decommission on-premises VMware hardware (data center exit, hardware end of life, lease termination) but have workloads that cannot quickly migrate to an alternative platform can move those workloads to VMware Cloud on AWS as a bridge. The workloads continue to run in vSphere. The on-premises hardware is decommissioned. The customer's longer-term goal is typically to migrate those workloads off VMware entirely, but the migration work continues in parallel with the on-premises hardware exit.

The decision between staying on VCF and migrating to an alternative comes down to a structured set of questions. The framework below produces a directional answer.

The first question is the renewal economic reality. What is the actual price the customer will pay at their next VCF renewal, and how does that compare to their prior VMware costs? If the increase is modest and operationally absorbable, the case for migration is weak. If the increase is significant relative to the alternative cost, the case for evaluation becomes strong.

The second question is the depth of VMware-specific dependencies. NSX-T deployment, SRM configuration, vRealize Suite usage, vCloud Director, PowerCLI automation, and specific application certifications create migration complexity. Environments with deep VMware-specific dependencies have higher migration cost and risk. Environments with primarily vanilla vSphere usage migrate more easily.

The third question is the existing storage architecture. Organizations with significant Pure FlashArray investment under active support have a clean path to disaggregated AHV. Organizations with vSAN have a clean path to Nutanix HCI. Organizations with other external storage need to evaluate compatibility with the target platform.

The fourth question is the operational team capability and capacity. Migration to an alternative platform requires team capacity for the migration project plus operational ramp-up time on the new platform. Teams already stretched on operational obligations may need managed services support to make the migration feasible.

The fifth question is the strategic technology direction. Organizations strategically committed to specific cloud platforms (Azure, AWS) may align their on-premises platform choice with that strategy. Organizations committed to multi-cloud or hybrid models may prioritize platforms with broader portability.

The sixth question is the timeline reality. VMware exits typically take 12 to 24 months for mid-size environments. Organizations with renewals that need to be addressed in the next 6 months may not have time for a full migration before the renewal. Some combination of negotiation and staged migration may be necessary.

The seventh question is the risk tolerance. Migration adds risk that the customer must accept. Some organizations are comfortable with the risk. Others prefer the lower-risk path of continued operation on the incumbent platform even at higher cost.

These seven questions almost always produce a clear directional answer. The specific platform choice within "alternative" (Nutanix, Azure Stack HCI, or some other option) follows from the directional answer plus the more specific technology fit questions.

IVI's engagement model is structured to support organizations at any stage of their VMware footprint evaluation, from early exploration through full migration execution.

For organizations earlier in their evaluation, IVI provides structured assessment engagements. The Infrastructure Modernization Assessment is the entry point. The assessment surfaces the specific characteristics of the customer's environment, maps them to the appropriate architectural recommendation, and produces actionable guidance about whether to remain on VCF, migrate to Nutanix, or pursue a hybrid path with Amazon VMware Cloud on AWS as a bridge.

For organizations actively planning a migration, IVI provides architectural design and migration planning. The output is a detailed migration plan including the target architecture, the wave-based migration sequence, the specific tools and procedures, and the timing.

For organizations executing a migration, IVI provides hands-on migration execution. IVI engineers run the Nutanix Move migration plans, manage the cutover windows, coordinate with application owners, and handle the operational discipline of the migration. The customer's team participates but is not solely responsible for the migration execution.

For organizations operating their Nutanix environment post-migration, IVI provides Aegis Managed Services. Aegis Performance Monitoring, Aegis Incident Response, and Aegis Lifecycle Management cover the operational scope of the Nutanix platform under a co-managed model.

The engagement model is sized to the customer's preferences. Organizations with strong in-house capabilities may engage IVI for specific phases (assessment, design, or specific migration support) while running the broader work with their own team. Organizations with limited in-house capacity may engage IVI for end-to-end migration and ongoing operations.

The honest framing is that IVI's commercial interest is in Nutanix migrations because that is where the active modernization work is happening. That said, IVI's recommendation is environment-specific. For organizations whose evaluation should land on remaining on VCF, IVI's professional advice is to remain on VCF rather than to force a migration that does not fit the environment. The Infrastructure Modernization Assessment is built to produce the right recommendation, not to default to one direction.