NVIDIA has introduced the DGX Station for Windows, a groundbreaking deskside system tailored to accommodate extensive AI workloads on machines operating with Microsoft’s Windows. This strategic launch signifies a shift towards integrating sophisticated AI development tools within the familiar Windows environments prevalent in large enterprises, moving away from the traditional reliance on Linux-based systems typically found in data centers.
Technical Specifications
The DGX Station for Windows is powered by the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, enabling it to execute AI models with up to 1 trillion parameters locally. This innovative workstation caters to a diverse array of professionals, including developers, researchers, engineers, designers, and data scientists, all of whom are engaged in the creation and deployment of AI agents.
Designed specifically for enterprise-level tasks, the DGX Station excels in model training, fine-tuning, inference, data science, and multi-agent development. Its capabilities extend to running hundreds of agents concurrently, seamlessly integrating them with business applications and workflow tools commonly utilized in Windows environments.
Innovative Features
A pivotal aspect of this launch is the NVIDIA OpenShell on Windows, a runtime environment for autonomous agents co-developed with Microsoft. OpenShell leverages advanced Windows security and containment features to isolate each agent within a secure sandbox, ensuring that policy controls remain outside the agent’s reach. This design addresses critical concerns regarding the interaction of autonomous software with enterprise systems, sensitive data, and user credentials. By implementing system-level controls, OpenShell prevents agents from overriding established policies or accessing protected information beyond the parameters set by administrators.
Chris Marriott, Vice President of Enterprise Platforms at NVIDIA, emphasized the growing demand for systems that facilitate direct connections between AI agents and the software utilized in workplaces. “As enterprises scale AI agents across their organizations, they need AI infrastructure that can connect directly to the applications and workflows that power their business,” Marriott stated. “DGX Station delivers supercomputing-class AI directly into Windows, where millions already design, engineer, research, and create every day.”
Collaboration with Microsoft
Microsoft has framed this initiative as a continuation of its longstanding collaboration with NVIDIA, aimed at enhancing Windows devices and larger computing systems. Pavan Davuluri, Executive Vice President of Windows + Devices at Microsoft, remarked, “For decades, Microsoft and NVIDIA have partnered to advance the most powerful computing platforms in the world. Today, we’re taking that collaboration to the next level, scaling the full power of Windows from thin-and-light PCs to data-center-class workstations with DGX Station powered by GB300. This unlocks a new class of AI performance on Windows, the platform enterprises trust for security, manageability, and compatibility.”
System Design and Enterprise Integration
The hardware architecture of the DGX Station combines a Blackwell Ultra GPU with a 72-core Grace CPU interconnected via NVIDIA NVLink-C2C. It boasts up to 748GB of coherent memory and can achieve up to 20 petaflops of FP4 performance. Additionally, it can be paired with an NVIDIA RTX PRO 6000 Blackwell Workstation GPU for enhanced visualization and simulation capabilities alongside AI processing. Networking is facilitated through an NVIDIA ConnectX-8 SuperNIC, supporting speeds of up to 800Gb/s, enabling multiple DGX Station systems to be interconnected for larger workloads.
NVIDIA has designed the system to integrate smoothly into existing enterprise management frameworks. For IT teams, the DGX Station for Windows extends established Windows security, compliance, and fleet management tools to a high-end AI workstation, while still accommodating Linux workloads through the Windows Subsystem for Linux.
Applications and Future Prospects
NVIDIA positions the DGX Station as a local AI compute node suitable for individual specialists or collaborative teams. Users can develop and run advanced AI agents on this system before transitioning workloads to larger GB300-based infrastructures in data centers or cloud environments. The company has outlined a comprehensive range of potential applications, including pretraining and fine-tuning large models, high-throughput inference, analytics on substantial datasets stored in system memory, and physical AI tasks that merge AI models with simulation and ray-traced visualization.
This launch reflects a broader trend in enterprise AI, moving from simple chatbot interfaces to sophisticated software agents capable of continuous operation, real-time task reasoning, and interaction with business systems. NVIDIA contends that deploying this work on a local Windows machine could diminish reliance on remote infrastructure for various development and deployment tasks, while keeping the software in closer proximity to the applications that employees utilize daily.
The DGX Station for Windows is anticipated to be available through prominent vendors such as ASUS, Dell Technologies, GIGABYTE, HP, MSI, and Supermicro.
