- Dell marks two years of its AI Factory with Nvidia, citing 4,000+ enterprise deployments
- The new Dell AI Data Platform features a no-code Data Orchestration Engine and Lightning File System
- Dell claims to be the first OEM to ship Nvidia's GB300 Grace Blackwell Ultra Desktop Superchip
NVIDIA GTC, SAN JOSE, CALIF. — Two years after launching its AI Factory with Nvidia, Dell Technologies is betting that enterprises have stopped asking whether AI works and started asking why it isn't working at scale. Dell says it has the answer.
"The gap between experimentation and production — that's where most organizations get stuck," Arthur Lewis, president of Dell Technologies Infrastructure Solutions Group said. "Here's what we've learned working with thousands of customers: you need a data platform that makes enterprise information AI-ready, infrastructure that scales from desktops to data centers, and solutions that accelerate your time to value. Those three things together are what moves AI from experiment to business driver."
To meet those goals, Dell announced enhancements spanning its AI data platform, hardware lineup and services portfolio, all framed around the argument that the era of AI as an experiment is over.
The announcements come as Dell claims its best enterprise AI quarter on record in Q4 2025, with more than 4,000 customers now deploying the Dell AI Factory with Nvidia, Dell said.
Strategic context and competitive pressure
The GTC announcements are emblematic of Dell's multi-year push to become a full-stack AI infrastructure company — not just a hardware vendor, but an integrated provider of compute, storage, networking, data management and deployment services. The AI Factory with Nvidia is foundational to that strategy, and the customer milestone gives Dell a proof point competitors can't easily replicate.
But the strategy faces real headwinds. HPE, Dell's most direct enterprise infrastructure rival, is pushing its own AI portfolio aggressively through the GreenLake platform and Cray HPC lines.
Cisco, meanwhile, came out with its own Secure AI Factory update at GTC, adding support for its new 102.4Tbps Cisco N9100 AI networking switch powered by Nvidia's Spectrum-6 Ethernet silicon, its Nexus Hyperfabric, an upgraded Hybrid Mesh Firewall that supports policy enforcement on DPUs, and AI Defense to enable implementation of agentic guardrails.
Elsewhere, Pure Storage and NetApp are competing sharply on AI-optimized storage. And the hyperscalers — AWS, Microsoft Azure and Google Cloud — are spending at a scale that forces Dell to keep remaking the case that on-premises infrastructure is more economical.
The data problem is the AI problem
The centerpiece of Dell's GTC news is the Dell AI Data Platform with Nvidia, a unified stack combining high-performance storage, modular data engines and Nvidia accelerated computing built to handle workloads ranging from retrieval-augmented generation and multimodal search to large-scale agentic workflows.
At the heart of the platform is a new Data Orchestration Engine, built on technology from Dell's recent acquisition of Dataloop. The no-code, low-code engine automatically discovers, labels, enriches and transforms structured, unstructured and multimodal data into governed, AI-ready datasets. It ships with a marketplace of more than 200 pre-built models, applications and templates, including Nvidia NIM microservices and AI Blueprints, so enterprises can deploy production-ready data workflows without building from scratch.
The platform integrates Nvidia RTX Pro Blackwell Server Edition GPUs directly into the data layer, using Nvidia CUDA-X libraries to deliver up to 3x faster SQL queries and 12x faster vector indexing. Time-to-first-token improves up to 19x compared to conventional approaches, according to Dell's internal testing.
The underlying logic is straightforward: with 83% of enterprise data residing on-premises — much of it in cold backup, dark storage or siloed formats — the bottleneck to performance is not the AI model. It's the data supply chain feeding it.
Storage built for AI scale
Dell also introduced Dell Lightning File System, which delivers up to 150 GB per second per rack, 2x greater throughput per rack unit compared to competing parallel file systems and up to 20x greater performance than traditional flash-only scale-out competitors, Dell said. The file system is designed to keep GPU-intensive workloads fed continuously rather than waiting on storage bottlenecks.
Complementing the file system is Dell Exascale Storage, a 3-in-1 platform combining file, object and parallel file system storage on a common hardware base, with planned connectivity up to 800 GbE and read performance up to 6 terabytes per second per rack.
From the desk to the data center
Dell is extending AI infrastructure reach all the way to the individual developer's workstation. The company announced it is the first OEM to ship Nvidia's GB300 Grace Blackwell Ultra Desktop Superchip with its Dell Pro Max desktop, delivering up to 20 petaflops of FP4 performance and 748 GB of coherent memory — bringing trillion-parameter-scale AI compute to a developer's desk without a cloud subscription or a data center request, Dell said.
On the server side, the PowerEdge XE9812 is a liquid-cooled server built around the Nvidia Vera Rubin NVL72 platform for massive training and inference workloads. The PowerEdge XE9880L, XE9882L and XE9885L round out the liquid-cooled lineup for organizations working within existing data center power and footprint constraints. The new PowerSwitch SN6000-series delivers Nvidia Spectrum-6 Ethernet with 1.6 terabits per second throughput and liquid-cooled co-packaged optics.
Closing the pilot-to-production gap
Hardware alone doesn't solve what Dell calls the gap between experimentation and production. A new modular architecture approach combines the Dell Automation Platform with Nvidia AI Enterprise software and pre-validated AI blueprints, compressing deployment timelines from months to weeks.
Dell also announced an Agentic AI Platform, built with Cohere's North and DataRobot, for designing, deploying and managing AI agents with governance and observability built in. Dell Accelerator Services for Agentic AI round out the portfolio with packaged professional services from initial validation to enterprise-wide rollout.
Two years after the launch of AI Factory with Nvidia, Dell has built a credible case that enterprise AI success is an infrastructure problem as much as a software one. The next two years will determine whether that case holds as cloud providers close the gap and rivals sharpen their own integrated plays.