VMware Explore 2025: Broadcom Transforms VCF into AI-Native Private Cloud PlatformVMware Explore 2025: Broadcom Transforms VCF into AI-Native Private Cloud PlatformVMware Explore 2025: Broadcom Transforms VCF into AI-Native Private Cloud Platform

VMware Explore 2025 reveals VCF 9.0 with integrated AI capabilities, automated security compliance, and streamlined developer workflows.

Picture of Sean Michael Kerner

Sean Michael Kerner , Contributor

August 28, 2025

5 Min Read

Hock Tan, Broadcom's president and CEO

Hock Tan, Broadcom's president and CEO, at VMware Explore 2025: "Since the acquisition [of VMware] two years ago, we rolled up our sleeves, did the tough engineering work, and the result today is VMware Cloud Foundation 9.0."

Over the last 25 years, VMware has built a business on the foundation of its virtualization technology.

The virtualization journey started with servers, then expanded to storage and networking, though those three domains largely remained siloed from each other. At the VMware Explore 2025 event, the company reiterated its promise to create a unified virtualization layer that enables virtual private clouds and the software-defined data center.

Recent surveys indicate that 70% of IT professionals are planning to return to on-premises deployments, driven by concerns over cost management, data sovereignty and security. At VMware Explore 2025, Broadcom positioned VMware Cloud Foundation (VCF) 9.0 as the technical answer to this "cloud reset," announcing an expanded platform that integrates AI capabilities, enhanced security, and developer productivity tools into a unified private cloud infrastructure.

Key Announcements from VMware Explore 2025

Key announcements include the following:

VMware Private AI Services integrated as standard components in VCF 9.0, including GPU monitoring, model stores, vector databases, and agent builders.
VCF Advanced Cyber Compliance service offering automated compliance enforcement, cyber recovery, and enhanced platform security.
Tanzu Data Intelligence data lakehouse platform for unified, low-latency access to multimodal data at scale.
Developer productivity enhancements including native S3 object storage, GitOps with Argo CD, and Istio service mesh integration.
Partnership expansions with Nvidia (Blackwell architecture), AMD (MI350 GPUs), and Canonical (Ubuntu integration).
Tanzu Platform 10.3 with granular AI model service plans, automated application modernization, and vulnerability dashboards.

"Well, last year I was here, we talked about private cloud being the future of the enterprise. Twelve months later, the future is here, and we have the data to back that up," said Hock Tan, Broadcom's president and CEO, opening the keynote. "Since the acquisition two years ago, we rolled up our sleeves, did the tough engineering work, and the result today is VMware Cloud Foundation 9.0: a rail to run all your application workloads with complete compute, networking, and storage tightly integrated."

AI Services Become Infrastructure Primitives

The most architecturally significant development is the transformation of AI capabilities from optional add-ons into core infrastructure services, which include:

Multi-Accelerator Model Runtime: The platform abstracts hardware dependencies across different GPU architectures. That includes Nvidia, Intel, and AMD.
Multi-Tenant AI Services: Single model instances can serve multiple business units or applications while maintaining strict data isolation.

"With multi-tenant models as a service, I can load a single copy of a model into one or more GPUs," Chris Wolf, global head of AI and Advanced Services for the VMware Cloud Foundation Division of Broadcom, explained during the keynote. "I can share that among multiple lines of business or multiple tenants, while keeping all of the data private. This will further lower your costs for AI services as you run them internally."

The platform also introduces VCF Intelligent Assistant as a practical implementation of these capabilities, providing AI-driven infrastructure troubleshooting that accesses Broadcom's technical knowledge base and documentation repositories.

Developer Platform Transformation

VCF 9.0 fundamentally alters the developer experience by addressing what Tan identified as a core friction point.

"Developers don't want to think about infrastructure," Tan said. "Modern apps run on containers, and developers just want to write code using their favorite DevOps tools."

VCF 9.0 fundamentally alters the developer experience by providing container-native services without requiring separate Kubernetes platform management. Several technical capabilities enable this approach:

Unified Storage Services: vSAN now provides native S3-compatible object storage alongside existing block and file interfaces. This eliminates the need for separate object storage infrastructure while enabling unified policy management across storage types. Developers can access S3 APIs directly from applications while IT maintains governance through vSAN policies.
GitOps-Driven Deployment: Argo CD integration enables Git repository-based infrastructure and application management. Developers define infrastructure requirements, security policies, and application configurations in YAML specifications stored in Git repositories. The platform automatically deploys and maintains desired state based on Git commits, providing audit trails and rollback capabilities.

Service Mesh Automation: Istio integration provides zero-trust networking, traffic management, and observability for containerized applications without manual service mesh deployment. This addresses the operational complexity typically associated with service mesh implementations while providing enterprise-grade security and monitoring capabilities.
Enterprise Ubuntu Foundation: VCF 9.0 now directly integrates with Canonical Ubuntu Linux. The Canonical partnership delivers enterprise-supported Ubuntu with automated security patching and "chiseled" container images that remove unnecessary libraries and dependencies. This approach reduces container attack surfaces while providing AI-ready images with pre-configured GPU drivers and machine learning frameworks.

[フレーム]

Data Lakehouse Architecture for AI Applications

Tanzu Data Intelligence addresses the fundamental challenge that up to 90% of enterprise data remains unstructured and inaccessible for AI and analytics applications. The platform implements a lakehouse architecture that combines data warehouse performance with data lake flexibility.

Unified Data Access Layer: The platform provides microsecond-latency queries across structured and unstructured data sources, scaling from terabytes to petabytes without performance degradation. Native vector search capabilities enable SQL queries and semantic similarity searches within the same query engine, eliminating the need for separate vector databases.
Federated Query Processing: Intelligent query optimization provides unified access across disparate data sources without requiring data replication. The query engine can access data residing in traditional databases, file systems, and cloud storage while maintaining governance and lineage tracking.
Streaming Data Integration: Real-time stream processing pipelines support event-driven architectures with automatic scaling based on data volume and processing requirements. This enables real-time AI inference and analytics on streaming data sources.

Tanzu Platform 10.3 introduces service publishing capabilities that enable developers to offer applications as reusable services through an internal marketplace. The platform also provides granular AI model quota management and cost allocation across business units or development teams.

"VCF 9.0 is the culmination of 25 years of VMware technology and innovation, and this is the platform for the future," Tan said. "We want to give you the best cloud platform in the world, and I mean that we want to make you a hero, the person who drives this huge impact in your organization."