VMware Explore: Broadcom CPO Answers Private AI Questions

## Private AI Cloud Strategy

Broadcom CPO Paul Turner discusses VMware Cloud Foundation's private AI direction at VMware Explore with TechTarget's Beth Pariseau. Enterprises want AI workloads on infrastructure they control rather than solely on public hyperscalers, driven by data sovereignty, regulatory pressure, and cost predictability.

## Hardware and NVIDIA Partnership

Turner highlights certified hardware support including NVIDIA H100 and Grace Blackwell ARM CPUs, plus ConnectX-7 networking offloads developed jointly with NVIDIA. The AI factory concept runs on VCF-certified stacks, maintaining a close though evolving vendor relationship as product branding shifts under Broadcom.

## Sovereign and Regional Deployments

Customers across Europe, Asia-Pacific, and Canada increasingly demand sovereign cloud deployments where data and models remain in-country. Turner positions "private AI cloud" alongside existing private cloud services, expecting banks, telcos, and insurers to extend on-premises patterns to model training and inference.

## Platform Engineering Implications

The conversation frames AI as another workload atop unified virtualization and Kubernetes management rather than a separate silo. Organizations must balance GPU capacity, networking, and lifecycle management within existing VCF operational models while meeting emerging compliance expectations for sensitive training data.