VMware Explore 2025 marked a decisive shift in enterprise cloud strategy, spotlighting how modern private cloud platforms are being re-architected to support the next generation of AI workloads, cyber resilience, and developer productivity. Broadcom’s announcements emphasized a unified AI-native infrastructure approach, tightly integrated security frameworks, and ecosystem-wide partnerships designed to help IT teams keep pace with growing complexity.
Below, we break down the key takeaways for CIOs, platform engineers, and IT operations leaders navigating the future of hybrid and private cloud.
The Shift From Infrastructure to Intelligent Cloud Fabric
Broadcom’s big takeaway was VMware Cloud Foundation (VCF) 9.0 is now more than a virtualization stack, it’s the operating system for the modern private cloud. With AI, security, and automation deeply embedded, VCF 9.0 positions itself as a unified control plane for compute, storage, networking, and increasingly, AI lifecycle management.
This evolution could address long-standing pain points in private infrastructure. IT leaders seeking to eliminate team silos, reduce operational complexity, and deliver faster developer experiences may now have a platform that unifies traditionally separate concerns. Features like live patching, global deduplication, memory tiering, and fast VPC creation aim to boost efficiency while minimizing downtime and technical debt.
By building in capabilities that once required a patchwork of third-party tools, Broadcom is altering what “cloud-like” means for on-prem deployments.
Private AI at the Core of Modernization
Broadcom formally embedded Private AI Services into the VCF 9.0 subscription, transforming VCF into an AI-native platform.
What’s Included:
- Multi-accelerator model runtime supporting AMD, NVIDIA, and CPU execution.
- Model Context Protocol (MCP) support for integrating structured/unstructured data.
- AI agent builder and vector database.
- Integrated model governance, API gateway, and data indexing tools.
Notably, these features are deployable on-prem and in air-gapped environments, with the goal of addressing both data sovereignty and AI readiness concerns.
“Customers don’t want to make a large investment in AI and be locked into a single provider,” said Broadcom engineers during the keynote. “VCF gives them the flexibility to pivot between models and accelerators.”
VCF Advanced Cyber Compliance
Security was another dominant theme at Explore 2025, with Broadcom unveiling VCF Advanced Cyber Compliance, a purpose-built service aimed at helping organizations maintain continuous regulatory compliance, even under active cyber threats.
Unlike traditional security offerings that focus on point solutions, this service promises to weave security, compliance, and resilience directly into the infrastructure layer. SaltStack-based enforcement could ensure real-time configuration compliance, while ransomware recovery capabilities may allow teams to restore workloads in isolated clean rooms.
Additional enhancements include hardened container images, confidential computing features, and extended threat detection via new vDefend and Network Detection and Response (NDR) sensors. The hopeful result is a platform that’s not only Zero Trust-ready, but purpose-built for the emerging threat landscape of AI-driven applications.
Intelligent Assist Gives AI-Driven Ops Support
VMware also introduced VCF Intelligent Assist, a GenAI-powered support assistant designed to help admins resolve issues quickly.
Features:
- Supports private and public LLMs.
- Points to relevant documentation, KBs, and remediation steps.
- Future roadmap includes automated resolution workflows.
This capability, paired with fine-grained vMotion for AI workloads, demonstrates VMware’s focus on AI operations (AIOps) and dynamic workload tuning
Ecosystem Expansion with Canonical, NVIDIA, AMD
A standout message at Explore 2025 was Broadcom’s continued investment in ecosystem openness and hardware flexibility. By deepening partnerships with Canonical, NVIDIA, and AMD, the company aims to ensure that VCF can meet customers wherever they are in their infrastructure journey.
Canonical’s Ubuntu OS is now tightly integrated with VCF, including support for chiseled containers, which are secure container images optimized for performance and minimal attack surface. For air-gapped environments, precompiled GPU drivers eliminate the need for runtime compilation, speeding up deployment while enhancing security.
Meanwhile, support for NVIDIA’s Blackwell B200 and AMD’s MI350 GPUs adds breadth to VCF’s hardware compatibility matrix. Combined with DirectPath I/O and advanced networking from BlueField-3 NICs, VCF becomes a flexible launchpad for AI inference, training, and high-performance computing workloads.
Tanzu’s Data Foundations
Modern applications need fast, low-latency access to multimodal data, and Broadcom responded by launching Tanzu Data Intelligence, a new lakehouse-style platform that integrates directly with VCF.
Built for speed and scale, Tanzu Data Intelligence supports both structured and unstructured data. It features streaming ingestion, vectorization for RAG pipelines, and unified access via Kubernetes-native APIs. The goal: simplify data movement and accelerate application delivery, especially in environments looking to bring AI to legacy data sets.
Tanzu Platform 10.3 also introduces role-based access controls, GitOps pipelines, and service mesh integration, which could give developers the tooling they need, and platform teams the control they require.
Real-World Adoption from Barclays, Walmart, and Grinnell Mutual
Broadcom showcased a range of customer deployments that demonstrate VCF’s scalability and relevance across sectors.
Barclays is leveraging VCF as the elastic infrastructure layer for its AI initiatives. The bank cited security, elasticity, and policy-based automation as major advantages that allow its teams to spin up secure AI environments on demand.
Walmart, meanwhile, is extending VCF across both internal data centers and retail edge environments. The move reflects a recommitment to VMware’s private cloud platform, especially for Kubernetes-based app development.
And in one of the most telling examples, Grinnell Mutual, a 17-person IT team, reported saving over $1M by moving from discrete storage to vSAN. The insurance company now runs mission-critical workloads across 19 states on VCF, proving that even lean teams can scale enterprise-grade infrastructure with the right platform.
What IT Leaders Should Watch
AI-Native Platform Wars Begin
VCF 9.0’s integration of model services, governance, and inference tooling signals a new chapter in private cloud competition. Broadcom is betting on embedded AI as the differentiator, positioning VCF not only as a virtualization or container platform, but as the AI operating layer inside the enterprise. This shift matters because IT decision makers increasingly want turnkey AI capability, not yet another integration project. Competitors will need to respond, either by deepening their own AI feature sets or partnering to deliver similar outcomes. Expect the “AI-native private cloud” to become a formalized category in 2026, with differentiation around developer tooling, data pipelines, and compliance readiness.
Cyber Compliance as a Service
Ransomware, regulation, and risk exposure are forcing compliance to become continuous and automated. VCF’s compliance enforcement and recovery workflows could become a model for others, as enterprises demand platforms that self-correct against drift and provide “push-button” recovery from ransomware. For IT leaders, this represents a strategic shift: compliance moves from being a manual, report-driven process to a built-in platform function. Over the next few years, buyers will likely prioritize platforms that can demonstrate provable compliance and incident recovery, especially in finance, healthcare, and government sectors where the cost of breaches is skyrocketing.
Hardware and Model Interoperability Wins
As organizations diversify their AI investments, the ability to mix GPUs, LLMs, and toolchains will define platform success. Broadcom’s embrace of open AI stacks gives it a clear edge, reducing the risk of lock-in at both the hardware and model layer. This flexibility not only lowers TCO but also allows enterprises to adapt quickly as new accelerators and models come to market. For CIOs and CTOs, the message is clear: choose platforms that can abstract hardware diversity and provide freedom of model choice. In an AI market evolving at breakneck speed, the ability to pivot without re-architecting infrastructure will separate winners from laggards.
The Bigger Picture
VMware Explore 2025 underscored that the private cloud conversation has shifted permanently. What once centered on virtualization efficiency now revolves around how well platforms can support AI-native workloads, enforce continuous compliance, and deliver flexibility across hardware and models. Broadcom used this event to position VMware Cloud Foundation as the strategic anchor for these priorities, offering embedded AI services, automated compliance, and expanded ecosystem partnerships in one unified platform. For IT leaders, the message from Explore 2025 is clear: the next phase of private cloud is about resilience, adaptability, and trust, and VMware intends to be the standard-bearer for that transformation.
