Supermicro's ARS-511GD-NB-LCC is a true desktop AI supercomputer purpose-built to power the next generation of AI development, research, and data science, with NVIDIA OpenShell support to enable safer development of autonomous, self-evolving agents.

Today at GTC 2026, Supermicro is thrilled to announce the Super AI Station (under product SKU ARS-511GD-NB-LCC), a first-of-its-kind platform that packages NVIDIA GB300 Grace Blackwell Ultra Superchip and NVIDIA ConnectX-8 Ethernet SuperNIC into a sleek and quiet liquid-cooled desktop tower form factor. This potent combination of exceptional compute performance and high-bandwidth, low-latency networking is perfect for developers looking to further push the boundaries of what’s possible with AI. We’re also working together on NVIDIA NemoClaw  — an open source stack that simplifies running OpenClaw always-on assistants, more safely, with a single command. As part of the NVIDIA Agent Toolkit, it installs the NVIDIA OpenShell runtime—a secure environment for running autonomous agents, and open source models like NVIDIA Nemotron. 

The Super AI Station is now available for purchase and interested parties are encouraged to visit the product landing page for details and visit Supermicro at GTC 2026, booth 1113.

Data-Center Grade Performance Meets Desktop Convenience

Positioned between consumer-grade PCIe-GPU based AI development desktops and high-end data-center grade AI servers, such as Supermicro’s HGX and NVL72 families of servers, the Super AI Station addresses several fundamental challenges that many organizations face while developing and researching AI. It provides the following benefits:

• Uncompromising hardware performance: This system combines 20 PFLOPS of uncompromising FP4 AI performance and 748GB of coherent memory, including 252GB HBM3e GPU memory and 496GB of CPU memory - making it an ideal platform for loading and acting upon both large models, up to 1 trillion parameters, and datasets. Additionally, the on-board dual-port 400GbE NVIDIA ConnectX-8 SuperNIC provides outstanding networking throughput, enabling massive data ingestion and seamless interfacing with other storage and development systems.
• Convenient form factor: The Super AI Station is a sleek, shared workspace-friendly on-prem solution. Equipped with a closed-loop liquid cooling system, the system is low noise, and its 1600W power supply can be powered using conventional power outlets. This plug-and-play design is ideal for teams looking to get started quickly without needing extensive infrastructure setup.
• Flexible deployment options: The Super AI Station can be deployed either as a 5U desktop tower, or within a rack.
• Tried-and-true architecture that powers production racks: The Super AI Station is an ideal platform for iterative development, experimentation and staging of candidate AI applications, as it is powered by the same data-center grade compute architecture that is used in production deployment environments. Developers can have the confidence that their code will scale to production without needing extensive modifications.
• Data privacy and security: The Super AI Station is a powerful on-prem platform that keeps proprietary data and intellectual property on-site without any cloud egress, satisfying compliance requirements.

Develop Powerful AI Agents On-Prem

With the rise of agentic AI and its natural evolution - autonomous, self-evolving systems, developers need powerful on-prem solutions that can handle compute-heavy agentic systems, along with robust software that emphasizes both data privacy and security. Today, NVIDIA also announced NVIDIA OpenShell – an open source runtime that enforces policy-based security, network and privacy guardrails that make autonomous, self-evolving agents safer to deploy. As the Super AI Station’s massive memory and compute performance make it the only deskside supercomputer capable of running frontier models up to 1 trillion parameters, it is the perfect platform to be paired with tools like OpenShell.

For enterprise production deployment, the Super AI Station can also support NVIDIA AI Enterprise, which provides microservices, frameworks, and libraries to streamline production AI development.

Key Technical Specifications: