NVIDIA RTX PRO Servers Redefine Enterprise Data Centers for the AI Era

Enterprise data centers are at a turning point. NVIDIA’s new RTX PRO Servers, powered by the RTX PRO 6000 Blackwell Server Edition GPU, are enabling companies to re-architect their infrastructure for AI without scrapping existing investments. Disney, Foxconn, Hitachi, Hyundai Motor Group, Lilly, SAP, and TSMC are among the first to adopt the systems - an early signal of how enterprise computing is shifting.

Justin Boitano, VP of Enterprise AI at NVIDIA, points to three forces driving this transition. "The world is moving from traditional computing to accelerated computing, enabling workloads from scientific research and quantum computing to enterprise applications like analytics, video processing, and computer graphics,” he said.

“At the same time, data centers are transforming into AI factories, purpose-built to train, deploy, and run AI models at scale. And third, AI is evolving from generating content to interacting with the physical world, powering robots and machines that can perceive, reason, and act.”

Together, these trends are reshaping how enterprises think about infrastructure. What were once IT cost centers are now becoming productivity engines.

A Platform Built for the Transition

For decades, enterprise servers were designed for predictable, low-power workloads, with racks filled by air-cooled systems constrained by power and density. “AI has broken that model,” Boitano noted. “It demands new levels of computational density. But introducing AI into existing IT infrastructure is a balancing act. Systems must integrate with legacy technology while being flexible and powerful enough to meet future demands. That’s where the RTX PRO Server comes in.”

The platform is built for enterprise deployment: air-cooled, PCIe-based, running on x86. It supports Windows, Linux, Kubernetes, hypervisors, and comes with enterprise-grade security and manageability. That means IT teams can run existing workloads while scaling new AI factories on the same infrastructure.

Early Enterprise Adoption

The momentum behind RTX PRO adoption is visible across industries.

"Across industries, companies are rethinking their data center strategies,” Boitano said. "They’re moving away from siloed infrastructure and embracing unified platforms that can power AI factories and a wide range of accelerated workloads. The RTX PRO Server is becoming the foundation of this new era of computing.”

RTX PRO Servers accelerate agentic AI, reasoning, simulation, and physical AI. NVIDIA reports up to 3× better price performance for the Llama Nemotron Super reasoning model compared to H100 GPUs; and up to 4× faster performance for digital twin, simulation, and synthetic data workflows compared to L40S GPUs. These improvements help enterprises scale advanced AI workloads while balancing cost and efficiency.

The New Foundation for Enterprise AI

RTX PRO Servers are supported by NVIDIA AI Enterprise software, Omniverse, and CUDA libraries, giving organizations a ready-made foundation for AI reasoning, simulation, and physical AI. Enterprises can buy or rent them through global OEMs like Cisco, Dell Technologies, HPE, Lenovo, and Supermicro, or access them in the cloud through CoreWeave, Google Cloud, and soon AWS, Nebius, and Vultr.

“This platform allows enterprises to transition from general-purpose clusters to AI factories without a full data center overhaul,” Boitano explained. “It’s how companies can reinvent IT operations and power their business into the next generation of AI.”

Turning Infrastructure into Channel Opportunity

Channel partners and service providers need to now rethink revenue models and managed services in light of the broad opportunities this platform opens up. With the shift toward sovereign AI and the adoption of a common infrastructure, the scope extends well beyond a single offering - covering virtual desktop delivery, design applications, scientific simulation, AI agent automation, and robotics.

Boitano answered ChannelE2E question on how channel partners and service providers should be thinking about building revenue models and managed services around the new NVIDIA offering.

Boitano emphasized, "We have such a broad ecosystem of service delivery partners, sometimes called NCPs, and we’re starting to see a major shift toward sovereign AI. This platform gives them a common infrastructure to support a wide range of use cases - from virtual desktops and design applications to scientific simulation, AI agents, and robotic simulation. For many managed service providers, these are entirely new markets. That’s why we’re seeing strong excitement around bringing this platform into their managed services infrastructure, alongside systems like HGX B200 or DB200, because of the breadth of workloads it can accelerate.”

By integrating technologies like the HGX B200 or DB200, partners can accelerate delivery, expand capabilities, and scale faster. Amdocs, for instance, is using the RTX PRO Server to build an agentic platform that improves both customer service and network operations for its global telco customers. This kind of deployment illustrates the tangible value of the platform in helping providers move beyond traditional service models. The goal now is to build adaptable, flexible service frameworks that can evolve alongside the expanding demands of AI and automation.