Broadcom Launches VMware AI Factory To Speed Enterprise AI Deployment

Broadcom’s VMware AI Factory automates AI infrastructure deployment and adds model governance, resource sharing and token-cost controls for enterprises.
Broadcom announced VMware AI Factory on August 31, introducing a software-defined foundation for VMware Private AI Cloud that is designed to speed up the deployment of enterprise AI workloads.
The company said VMware AI Factory can reduce the time required to move from bare-metal server deployment to serving an AI model from weeks to hours. The platform automates hardware provisioning, software stack enablement and lifecycle management through VMware Cloud Foundation (VCF).
The announcement comes as enterprises move more AI workloads into production, making the cost and performance of inference an increasingly important infrastructure concern.
Broadcom Targets AI Deployment And Token Economics
Broadcom said VMware AI Factory is designed to help enterprises run AI applications alongside private enterprise data while maintaining control over infrastructure and model deployment.
The platform pools and shares graphics processing unit (GPU) resources across organizations, allowing multiple models to use shared hardware rather than dedicating infrastructure to individual workloads. It also provides a unified model gallery for deploying and managing model inference and Retrieval-Augmented Generation (RAG) workflows across virtual machines, containers and GPU resources.
The system includes observability for token throughput, latency, and compute and memory utilization. Broadcom is also adding private AI services including multi-tenant model sharing, an AI Gateway and secure AI sandboxes.
The AI Gateway is designed to provide a single interface for model consumption across on-premises and cloud environments. It includes prompt routing, token and usage rate limits and application authorization, putting model selection and usage controls at the infrastructure layer. A similar focus on model routing and AI economics is emerging elsewhere in enterprise technology.
Paul Turner, Chief Product Officer, VMware Cloud Foundation Division, Broadcom, said enterprises wanted to run AI where their data lives but faced a slow and complex path from infrastructure to production models. He said AI Factory was designed to automate infrastructure deployment and lifecycle management while giving customers flexibility over hardware and models.
Broadcom Expands Hardware And Infrastructure Options
Broadcom is also partnering with MetalSoft to automate heterogeneous bare-metal infrastructure for VCF. The companies said the integration can reduce physical server provisioning from weeks to minutes and allow IT teams to provision or repave servers from multiple vendors through the VCF management console.
VMware AI Factory combines VCF with certified Dell PowerEdge servers and VCF AI ReadyNodes from Cisco, Lenovo, Supermicro and other vendors. Broadcom is also collaborating with AMD to pair VCF with AMD Instinct GPUs and the open AMD ROCm software ecosystem.
Broadcom said zero-touch provisioning will automate deployment across the stack, including vSphere, vSAN, Kubernetes and the AMD GPU operator. AMD has separately positioned ROCm as a software foundation for moving workloads from Instinct infrastructure into production inference and training.
The company said VCF customers can run more than 150 open-source and commercial AI models through the platform, including Nemotron 3, Gemma 4, cotomi, Qwen and GLM 5.2. Broadcom said the approach is intended to give enterprises a production path for running AI models on-premises while maintaining data sovereignty and control over infrastructure costs.
The announcement expands Broadcom’s AI infrastructure strategy into the software and operations layer, alongside its existing semiconductor and networking businesses.
Key Takeaways
- Broadcom's VMware AI Factory automates and accelerates enterprise AI infrastructure deployment from weeks to hours.
- The platform provides model governance, resource sharing, and token-cost controls for AI workloads.
- VMware AI Factory enables enterprises to run AI applications privately with data while maintaining infrastructure control.