Nscale brings up its first NVIDIA Vera Rubin NVL72 AI supercomputers

November 18, 2026

London, U.K., 18 November 2026 – Nscale has brought online its first NVIDIA Vera Rubin NVL72 systems, using the newest rack-scale supercomputers for artificial intelligence to launch a global rollout expected in the coming months.

The systems are Dell PowerEdge XE9812 servers built on the NVIDIA Vera Rubin NVL72 design, and give companies that develop and run AI more of the computing power they need. Customers will use them to train and run large AI models that reason through complex problems step by step.

Each Vera Rubin NVL72 rack holds 72 NVIDIA Rubin graphics processing units, or GPUs, the chips that perform the heavy maths behind AI, along with 36 NVIDIA Vera CPUs, memory, and high-speed connections. The components are designed to operate as one AI supercomputer.

With this launch, Nscale became the first in the industry to deploy a new Dell networking rack that is assembled and tested at the factory, complete with switches, power, and liquid cooling, and shipped fully populated. That eliminates network rack assembly on site and speeds deployment.

“Nscale’s deployment marks an industry first for a factory built, fully integrated, liquid-cooled networking rack. Dell PowerRack integrates and validates compute, networking, power and cooling in the factory, reducing the work required onsite and helping Nscale bring its high performance compute infrastructure into production faster. It’s an important step in delivering complete AI infrastructure at rack scale.” - Ihab Tarazi, CTO, Compute and Networking, Infrastructure Solutions Group, Dell Technologies

The bigger challenge is getting many GPUs to work together on a single job without sitting idle while they wait for data. Nscale links the GPUs with two layers of networking: NVIDIA NVLink high-speed connections inside each rack and a purpose-designed, standards-based version of Ethernet, the common networking standard, between racks.

Nscale’s Fleet Operations platform manages each system from setup and burn-in testing through repairs and retirement, and monitors the entire fleet in real time to catch problems early. Customers can see the signals they need to run on the platform, including what capacity is available and what is under maintenance.

Demand for such systems is rising as AI evolves. The most advanced models now carry out multistep tasks on a user’s behalf, requiring far more computing power than earlier AI applications.

“The next generation of AI factories requires tightly integrated compute and networking to support reasoning and agentic AI at scale,” said Dion Harris, senior director of HPC and AI hyperscale infrastructure solutions at NVIDIA. “With NVIDIA Vera Rubin NVL72, Nscale is expanding the infrastructure customers need to train more capable models and scale inference for advanced AI services.”
“Demand for AI compute is carrying our customers into a new age of reasoning and agentic workloads, and NVIDIA Vera Rubin is the platform built to enable it,” said Nidhi Chappell, President of AI Infrastructure, Nscale. “We have brought up our first NVIDIA Vera Rubin NVL72 racks, and over the coming months we expect to continue a global rollout as a single production environment, managed by the same fleet automation and observability as the rest of our cloud. Because we operate the full stack from ground to cloud, we control that timeline, and customers get next-generation compute at the scale and efficiency frontier workloads demand.”

Additional systems are expected to come online in the months ahead. Customers can contact the company to place orders or discuss timing.

‍

‍

‍

About Nscale

Nscale Limited (Nscale) is a full-stack AI cloud platform. We bring together software, compute, and power in a vertically integrated offering – from a unified cloud platform for running AI training and inference, to the data centers and low-cost power that make it possible. Our mission is to build the engine of superintelligence and enhance access to the benefits of advanced AI for enterprises, governments, and the communities that depend on them.

‍

Press contact: press@nscale.com

IR contact: ir@nscale.com

Stay up to date with Nscale

By submitting you agree to receive Nscale emails & accept our Terms & Privacy Policy.