Dell marks two years of its AI Factory with Nvidia at GTC 2026, announcing new data, infrastructure and storage tools to help enterprises move AI from experimentation to scale.
Dell Technologies used Nvidia’s GTC 2026 conference on Sunday to mark the second anniversary of its AI Factory with Nvidia partnership and announce a sweeping set of updates across data, infrastructure and services — all aimed at solving what the company describes as the defining problem in enterprise AI: not building models, but making them work at scale in production environments.
The announcements come as unclear return on investment remains the top obstacle preventing AI deployments from moving beyond the pilot phase. Dell said more than 4,000 customers are now deploying the AI Factory, with early adopters reporting up to 2.6 times return on investment within the first year.
“Two years ago, enterprises were asking how to access AI technology,” said Michael Dell, the company’s chairman and chief executive. “Today, they’re asking how to make their data AI-ready, how to operationalize AI at scale and how to prove ROI.”
Jensen Huang, Nvidia’s founder and chief executive, framed the partnership in sweeping terms. “AI infrastructure is being built everywhere — every company will be powered by it, every country will build it,” he said. “Dell Technologies delivers integrated data platforms, scalable infrastructure and deployment expertise, with Nvidia at the core.”
The Data Problem
The centerpiece of Sunday’s announcement is an expanded Dell AI Data Platform with Nvidia, designed to address what both companies describe as the most persistent bottleneck in enterprise AI: data that is too slow, too siloed or too unstructured to be useful.
A new Data Orchestration Engine, built on technology from Dell’s recent acquisition of Dataloop, automates the complete AI data lifecycle — discovering, labeling, enriching and transforming structured, unstructured and multimedia data into governed, AI-ready datasets without requiring custom engineering work. An accompanying marketplace offers more than 200 pre-built models, applications and templates, allowing organizations to deploy production-ready data workflows without building them from scratch.
Dell also announced GPU-accelerated data processing capabilities, combining Nvidia’s CUDA-X libraries with Dell’s storage infrastructure to deliver up to 12 times faster vector indexing, three times faster data processing and 19 times faster time-to-first-token compared with traditional computing approaches, the company said. A new AI Assistant within the Dell Data Analytics Engine brings a natural language interface to SQL analytics, allowing business users to query and visualize governed data without specialized technical knowledge.
Also Read: Amazon’s Layoffs Aren’t a Warning. They’re a Preview.
Storage Built for AI at Scale
As enterprises move from experimentation to production, Dell argues that storage has become the critical constraint — traditional architectures slow as they scale, leaving expensive GPU resources idle while waiting for data.
To address this, Dell announced two new storage systems. Dell Lightning File System, which the company describes as the world’s fastest parallel file system, delivers up to 150 gigabytes per second per rack for AI training and inference environments. Dell Exascale Storage, a three-in-one system combining file, object and parallel file storage on a single hardware platform, targets the most demanding AI and high-performance computing environments, delivering read performance of up to six terabytes per second per rack.
Both systems support Nvidia’s latest infrastructure, including CMX context memory storage — a capability that allows AI agents to offload memory from GPU hardware to shared network storage, enabling systems to maintain context across extended interactions without exhausting GPU resources.
New Hardware, From Desktop to Data Center
Dell also announced a broad refresh of its hardware portfolio spanning every scale of AI deployment. At the desktop level, the company said it is the first original equipment manufacturer to ship a system powered by Nvidia’s GB300 Grace Blackwell Ultra Desktop Superchip, delivering up to 20 petaFLOPS of computing performance for developing and deploying autonomous AI agents. Dell Pro Precision workstations — available in tower configurations with up to five Nvidia RTX PRO Blackwell GPUs and in mobile configurations — target AI developers and data scientists.
For data center deployments, Dell announced the PowerEdge XE9812, its flagship liquid-cooled server built on Nvidia’s Vera Rubin NVL72 platform for large-scale AI training and inference, expected to be globally available in the second half of 2026. Several additional liquid-cooled server configurations were announced for organizations deploying AI within existing data center power and space constraints.
In a notable first, Dell said it has integrated Nvidia NVQLink with CUDA-Q across its PowerEdge server line, making it the first hardware maker to combine classical and quantum computing capabilities on a single platform — a capability the company said could accelerate research in drug development and materials science.
Also Read: Cutting Through the Noise of SaaS Buying
Services to Close the Skills Gap
Alongside the hardware and software announcements, Dell introduced updated AI solutions and services designed to compress the time between investment and measurable outcome. New agentic AI platform offerings, developed in collaboration with Cohere, DataRobot and ClearML, allow enterprises to deploy and manage AI agents with built-in orchestration, governance and monitoring. Dell Accelerator Services for Agentic AI provide packaged consulting support for organizations at any stage of AI deployment, from initial experimentation to enterprise-wide integration.
