Announcing General Availability of VMware Cloud Foundation 9.1.1

Coming on the heels of a very successful VMware Explore in Vegas and the VMware Cloud Foundation (VCF) 9.1 launch in May, we’re excited to announce the general availability of VCF 9.1.1. This release builds on VCF 9.1 with tougher security, vSAN Object Storage (tech preview, previously announced) and new capabilities designed to make your private cloud the best place to run, govern, and secure AI workloads — while lowering TCO and reducing operational complexity.

Let’s look at what’s new.

As organizations deploy more AI models across business units, they face a familiar tension: teams want autonomy to consume models, but MLOps and platform engineers must maintain data privacy and governance. Running separate model stacks per tenant drives up cost and operational sprawl.

Multi-Tenant Model Sharing solves this by enabling the secure sharing of AI models between tenants or separate lines of businesses while maintaining data privacy. Teams can consume shared models without exposing sensitive training data or creating isolated infrastructure silos.

Figure 1: Multi-Tenant Model Sharing in VMware Cloud Foundation

The operational benefits include:

Lower TCO through shared model infrastructure instead of redundant per-tenant deployments
Simplified operational complexity with a single management plane for model access and policy
Data privacy preserved through tenant-isolated access controls

This is what it looks like when a platform is built for AI from the ground up, not bolted on after the fact.

AI Assistant for VCF: Diagnostics and Management Packs Builder

Infrastructure teams spend significant time on root-cause analysis and on troubleshooting where issues are in the adjacent infrastructure. As environments scale to hundreds of hosts and Kubernetes clusters, that effort compounds.

The AI Assistant for VCF assists with diagnostics and builds management packs. AI diagnostics accelerates mean time to repair (MTTR) by helping teams surface root causes faster. Using AI, users can resolve issues—such as ESX CPU and memory contention or other diagnostic anomalies—without needing specialized skills. Another example of troubleshooting with the AI Assistant is analyzing multiple vMotion errors together to pinpoint why vMotion is failing across a vCenter environment.

The AI Assistant for VCF helps customers create third-party integrations with a content builder by leveraging APIs to create management packs and content packs that would otherwise require manual work.

By combining AI-driven diagnostics with deeper third-party visibility, modern private clouds eliminate operational bottlenecks so infrastructure teams can stay proactive rather than reactive.

New GitOps Service [Tech Preview]: Argo CD, Native in VCF

Platform engineering teams increasingly rely on GitOps to automate deployment and lifecycle management. But integrating GitOps tooling has historically meant operating outside the VCF management framework, adding toolchain complexity and support ambiguity.

Our new GitOps Service (releasing as a tech preview) adds Argo CD as a native GitOps service within the VCF service management framework. This unifies deployment and lifecycle management directly from the VCF Automation UI, bringing GitOps workflows into the same operational plane as the rest of your infrastructure. By reducing toolchain sprawl, teams improve their operational efficiency and ROI. The GitOps Service also safeguards pipeline access using VCF Automation-integrated OIDC authentication.

We’re offering platform engineering teams early access via Tech Preview to partner with us as we refine the feature for platform engineering teams at scale.

As private cloud environments grow, network complexity grows with them. Networking teams struggle with complex configuration, performance limitations, and associated troubleshooting friction.

EVPN Architectural Enhancements: Simpler, Scalable Networking

VCF 9.1.1 introduces key EVPN architectural enhancements, including an optimized datapath between VCF transit gateways and the physical fabric for East-West traffic. Direct VXLAN tunnels between the transit gateway and leaf switches optimize datapath performance for East-West traffic. In addition, networking services such as NAT, NSX Load Balancing, AVI Load Balancing, and DHCP relay are now available for the EVPN VXLAN distributed connectivity mode. Together, these enhancements enable critical multi-tenant and external connectivity and deliver better performance, simpler configuration, and easier troubleshooting of EVPN environments.

This enhanced interoperability with the physical fabric means network teams can design and operate larger, more flexible private cloud networks without the complexity that has traditionally slowed them down.

VMware vSphere Kubernetes Service 3.7 Add-on Management Framework: Confidence to Scale

As Kubernetes environments grow, the ecosystem of add-on tools — networking, observability, security, storage — grows with them. The real challenge hits on Day 2: managing add-on lifecycles and troubleshooting issues as they arise. Unclear support ownership makes it worse — not knowing who to call when something breaks. The vSphere Kubernetes Service (VKS) 3.7 Add-on Management Framework delivers predictable support with transparent ownership and gives customers choice and flexibility in adopting ecosystem tools. Clear support paths across every tier reduce time to resolution, giving teams the confidence to scale and the speed to innovate.

Take the next step:

VCF 9.1.1 is available now. If you’re already running VCF 9.1, this release continues our cadence of secure enhancements designed to keep your private cloud at the forefront of AI infrastructure. If you’re evaluating VCF, there’s never been a better time to make the move.

Learn More:

VCF 9.1.1 Release Notes (Techdocs)

From Bottleneck to Breakthrough: Centralizing GitOps at Enterprise Scale with VCF 9.1.1

New AI and Kubernetes Private Cloud Operations Capabilities in VMware Cloud Foundation 9.1.1

Multi-Tenant Model Sharing: Share AI Models Securely