Dell Expands AI Factory With Nvidia at Supercomputing 2026

Dell Technologies has announced a broad set of upgrades to its AI Factory with Nvidia, using the Supercomputing 2026 conference in Chicago on Monday, 16 November 2026, to pitch enterprises on moving artificial intelligence from pilot projects into full production. The updates cover data preparation, storage, infrastructure management and ready-made AI solutions, and Dell says the platform now spans more than 420 offerings and serves more than 6,500 customers.
The announcement, made at the annual high-performance computing gathering known as SC26, builds on the Dell AI Factory with Nvidia, the joint hardware and software line the two companies launched to give businesses a complete foundation for running AI systems on infrastructure they own and control.
Data first: a gateway for AI agents
The centrepiece of the data announcements is the Agentic Data Gateway, part of the Dell AI Data Platform. It is designed to give every AI agent inside a company a single, governed route to enterprise data, so teams connect their data once instead of building a separate link for each agent. Dell is also adding new data orchestration and storage engine capabilities to help customers prepare data wherever it lives.
Software-defined PowerScale running on Dell Exascale has achieved Nvidia-Certified Storage validation at the Foundation level, giving organisations a validated platform for AI-accelerated data access in large graphics processor environments. The AI Data Platform advancements will be available from December 2026.
Liquid-cooled storage and validated servers
On storage, Dell is expanding its Exascale portfolio with what it describes as the world’s first direct-liquid-cooled storage for rack-scale AI. The two-unit platform supports up to 32 NVMe drives and is expected in early 2027, alongside a one-unit option with 67 percent higher drive density per rack unit.
Dell has also validated its PowerEdge XE9785 servers, fitted with Nvidia HGX B300 graphics processors, for the AI Factory’s modular architecture. Those servers are available now, as is Dell-supported Nvidia AI Enterprise software for the platform. PowerStore validation for the AI Factory will follow in the fourth quarter of calendar 2026.
Managing up to 100,000 devices
The management software is the boldest scale claim in the announcement. Dell OpenManage Rack Scale, the next generation of its OpenManage platform, deploys, manages, observes and automates physical infrastructure from single servers to complete racks, at a scale of up to 100,000 devices — up to 12 times the previous generation, according to Dell. It is available today, as is Dell Fabric Manager.
Dell is also releasing validated blueprints for common AI workloads, including an agentic AI platform built with DataRobot and a code assistant reference architecture built with OpenAI. Several further blueprints are due in the first quarter of 2027.
What it means for enterprise AI
The package is aimed squarely at a familiar enterprise problem: companies have bought powerful AI hardware but struggle to feed it clean, governed data and to manage it at scale. Dell’s answer is to sell the whole stack — data, compute, networking, software and services — as one tested system, with engineering-led delivery through its new AI Factory Engineers model.
Buyers should still read the small print. Several of the headline items, including the liquid-cooled storage and some blueprints, do not arrive until early 2027, and Dell’s performance and scale figures are the company’s own. Even so, the timing — at the largest supercomputing event of the year — underlines how central AI infrastructure has become to Dell’s business.
Source: Dell Technologies; see also Blocks and Files.