Broadcom launches VMware Cloud Foundation 9.1 for AI
Wed, 6th May 2026 (Today)
Broadcom has launched VMware Cloud Foundation 9.1, an update aimed at companies running AI workloads in private cloud environments.
The release adds support for inference, agentic AI applications, containers and virtual machines on a single infrastructure stack, while extending hardware support across AMD, Intel and NVIDIA systems. It also introduces integrated security controls, observability tools for AI workloads and management changes intended to reduce infrastructure and operating costs.
The announcement comes as businesses weigh where to place production AI systems. Broadcom cited preview findings from its Private Cloud Outlook 2026 research showing that 56% of organisations are running or planning production inference in private cloud environments, while 41% are using or planning public cloud for the same task, down year on year.
The same research pointed to pressure on budgets and governance. Broadcom said 62% of IT leaders were very or extremely concerned about generative AI infrastructure costs, while 36% said AI was creating new requirements for data protection, privacy, security controls and risk management.
Cost focus
Broadcom positioned the update around lower spending on servers, storage and Kubernetes operations. It said VMware Cloud Foundation 9.1 can cut server costs by up to 40% through memory tiering in clusters running a mix of AI and non-AI workloads, lower storage total cost of ownership by up to 39% through compression and deduplication, and reduce Kubernetes operational costs by up to 46% for AI workloads at scale.
The platform can also handle larger deployments with faster maintenance cycles. Broadcom said fleet management capacity reaches up to 5,000 hosts and that cluster upgrades are up to four times faster.
The changes reflect a broader effort to make private cloud infrastructure more suitable for AI deployments moving beyond pilot projects. The software is designed to let companies run both CPU-heavy agentic workloads and GPU-based inference on the same platform, rather than splitting them across separate systems.
Broadcom said the release increases Kubernetes cluster scale by 2.6 times and cuts deployment times by 70%, with shorter upgrade windows than earlier preview versions. It has also added live application blueprints to capture multi-virtual-machine application stacks as reusable templates for development, test and production environments.
Security push
Security is a central part of the release. VMware Cloud Foundation 9.1 adds on-premises ransomware recovery with isolated recovery environments and validation tools, continuous compliance enforcement, and live patching that supports up to 80% of use cases without host evacuation or maintenance windows, Broadcom said.
The release also extends zero-trust segmentation and distributed intrusion detection and prevention to Kubernetes AI workloads. Broadcom said the platform includes centralised governance and policy controls to help organisations manage data sovereignty and regulatory requirements.
Hardware support is another area of emphasis. VMware Cloud Foundation 9.1 supports multi-accelerator environments spanning AMD and NVIDIA GPUs, along with AMD and Intel CPU platforms. Broadcom also highlighted support for NVIDIA ConnectX-7 network interface cards and NVIDIA BlueField-3 with Enhanced DirectPath I/O, which it said enables high-speed multi-host AI model training and data transfer.
Network interoperability with Arista systems is included through EVPN and VXLAN, while load balancing and security functions can be provided through VMware Avi Load Balancer and VMware vDefend. Broadcom said this reduces the need for separate hardware appliances for AI inference endpoints and related applications.
Krish Prasad, senior vice president and general manager of Broadcom's VMware Cloud Foundation division, outlined the company's pitch for the release.
"As more enterprises turn to AI for driving competitive advantage, they face three critical challenges: data and IP privacy concerns, surging infrastructure costs, and their readiness for the world of agentic AI," said Prasad. "VCF 9.1 is a single unified platform that addresses all three and delivers one of the most advanced infrastructures for Private AI. We enable zero-trust security for AI, reduce costs through intelligent infrastructure optimization and hardware choice, and provide the flexibility to run both agentic workflows and accelerated inferencing on the same platform."
Customers and partners linked the release to practical deployment concerns including cost control, security and infrastructure choice. Broadcom included comments from users in publishing and emergency services, as well as from chipmakers and infrastructure partners.
"Analyzing years of news archives in the public cloud is cost-prohibitive, with unpredictable pricing that makes AI projects difficult to plan," said V V Jacob, senior general manager, systems, at Malayala Manorama Co Ltd. "By deploying VCF Private AI Services on our existing VMware Cloud Foundation infrastructure, we will run AI-powered content summarization, heading generation, and editorial assistance directly on our private cloud. We believe this will give us the privacy and security essential for protecting editorial sources while delivering the cost predictability that on-premises private cloud infrastructure provides."
"By unifying our VMs and containers on VMware Cloud Foundation, we've achieved greater operational efficiency and raised the overall availability," said Hopfgartner. "VMware vSphere Kubernetes Service, as the built-in Kubernetes runtime of VCF, empowers our operations team to easily deploy, scale, and manage our most critical applications."
"Enterprises need infrastructure that delivers breakthrough AI performance while maintaining data sovereignty and control," said Fanelli. "Our collaboration with Broadcom brings NVIDIA Blackwell architecture-including RTX Pro Servers equipped with BlueField-3 and the NVIDIA Blackwell HGX platform-along with high-speed DirectPath I/O to VMware Cloud Foundation. This enables organizations to deploy private AI with the same performance they expect from public cloud, but with complete control over their models and data. This collaboration addresses the reality that production AI requires both extraordinary compute power and enterprise-grade governance."