New NVIDIA RTX PRO Servers to Accelerate AI Factories

At COMPUTEX this week, NVIDIA announced its NVIDIA RTX PRO Servers and a new NVIDIA Enterprise AI Factory validated design for building data centers that, according to the company, will "deliver universal acceleration for AI, design, engineering and business applications."

Built with NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, RTX PRO Servers extend the performance and energy efficiency of the NVIDIA Blackwell architecture to data centers that can run enterprise workloads, which the company says will drive the shift from CPU-based systems to GPU-accelerated infrastructure.

According to NVIDIA, using the NVIDIA Enterprise AI Factory validated design, partners are building a new class of on-premises infrastructure — featuring RTX PRO Servers, NVIDIA Spectrum-X Ethernet networking, NVIDIA BlueField DPUs, NVIDIA-Certified Storage systems and NVIDIA AI Enterprise software — to accelerate product design and engineering simulation applications, as well as AI-enabled business systems and teams of digital AI agents. 

“AI is revolutionizing every industry — every company will build or rent AI factories to run their businesses and power the intelligence of their products,” said Jensen Huang, founder and CEO of NVIDIA. “With our global partner ecosystem, we’re helping enterprises infuse AI into their workforce, automate their factories and build AI-native products.”

Cadence, Foxconn and Lilly are among the first NVIDIA partners planning to build AI factories using the NVIDIA Enterprise AI Factory validated design with RTX PRO Servers to advance their operations.

Foxconn is building an AI factory to advance global semiconductor and electronics manufacturing with NVIDIA-accelerated IT infrastructure, as well as its smart electric vehicles, factory digital twins, healthcare and robotics applications.

“Foxconn is harnessing the performance of NVIDIA Blackwell to build AI infrastructure that will transform every facet of electronics manufacturing,” said Young Liu, chairman of Foxconn. “Through our close collaboration with NVIDIA, we will accelerate the integration of AI across our global operations and deliver smarter electronics for the world.”

The RTX PRO 6000 Blackwell is a universal data center GPU for powering AI factories and accelerating demanding enterprise AI workloads, from multimodal AI inference and physical AI to design, scientific computing, graphics and video applications, the company says.

NVIDIA RTX PRO Servers from global system partners can support up to eight NVIDIA RTX PRO 6000 Blackwell GPUs, including NVIDIA BlueField-3 DPUs and NVIDIA ConnectX-8 SuperNICs with built-in PCIe Gen 6 switches.

Enterprises can accelerate AI and data science workloads on RTX PRO Servers with the NVIDIA AI Enterprise software platform. Developers can optimize AI inference performance and agent accuracy with NVIDIA NIM and NeMo microservices, and use the latest NVIDIA AI Blueprints for digital humans and AI query engines.

With powerful NVIDIA RTX graphics and AI capabilities, NVIDIA RTX PRO Servers deliver exceptional performance for industrial digital twin and robotics learning and simulation workflows developed on the NVIDIA Omniverse platform.

Using the NVIDIA Enterprise AI Factory validated design — based on recommended hardware configurations from NVIDIA Enterprise Reference Architectures — NVIDIA partners will build Blackwell AI factories featuring NVIDIA-Certified Servers, NVIDIA Spectrum-X, NVIDIA BlueField, NVIDIA-Certified Storage and NVIDIA AI Enterprise software.

The full-stack, validated design provides AI software stack recommendations to help enterprise customers build and operate on-premises AI factories. It offers guidance for scaling an enterprise AI factory with RTX PRO Servers, including deployment best practices, to help organizations meet their growing AI business needs efficiently and reliably.

Customers can also architect their Blackwell AI factories with NVIDIA HGX B200 systems for large-scale AI workloads.

NVIDIA ecosystem partners are building products, software and services to speed the enterprise IT shift to accelerated AI factory infrastructure.

Global system makers Cisco, Dell Technologies, Hewlett Packard Enterprise and Lenovo will offer full-stack solutions with NVIDIA RTX PRO Servers and NVIDIA AI Enterprise software using the NVIDIA Enterprise AI Factory validated design.

Data center system partners including Advantech, ASRock Rack, ASUS, Compal, Foxconn, GIGABYTE, Inventec, MiTAC Computing, MSI, Pegatron, Quanta Cloud Technology, Supermicro, Wistron and Wiwynn will also be offering NVIDIA RTX PRO Servers.

Blackwell AI factories are ideal for accelerating workloads across a broad range of popular enterprise software platforms, including those from Ansys, Cadence, CrowdStrike, Elastic, Red Hat, Siemens and Synopsys. 

Enterprises building AI factories with NVIDIA RTX PRO 6000 Blackwell GPUs can deploy NVIDIA-Certified Storage from partners including DDN, Dell, HPE, Hitachi Vantara, IBM, NetApp, Nutanix, Pure Storage, VAST Data and WEKA to support a broad range of workloads.

Sources: Press materials received from the company and additional information gleaned from the company’s website.