Dell says its AI Factory with NVIDIA has over 6,500 customers, and is adding deskside systems, security tools, validated partner solutions and new servers to help companies take autonomous AI agents from pilots into production.
Â
Dell Technologies has announced a wide set of additions to its Dell AI Factory with NVIDIA, covering partner software, security, desktop machines and data centre hardware. The aim is to help businesses move agentic AI, meaning software agents that act on their own, out of the experimental stage.
The company said the platform now has more than 6,500 customers, up from 5,000 last quarter, based on its own analysis of order data from August 2026.
Why Dell says it matters
Dell cited industry forecasts that agentic AI will become the largest enterprise AI workload by 2028 and will grow nearly three times faster than generative AI by 2030. It also pointed to projections that inference costs will fall by more than 90% by 2030 while agents drive a 24-fold rise in token consumption.
Many companies still struggle to justify spending. Dell cited a survey showing unclear returns are the biggest obstacle for 59% of enterprise AI deployments, and another saying 83% rank end-to-end vendor support as their top priority. Dell argued that owning infrastructure gives firms more control, and claimed its AI Factory can be up to 90% cheaper than public cloud over two years. That figure comes from analysis cited by Dell in its release.
“Enterprises used to think about AI infrastructure as an IT budget question, but that’s changed. AI infrastructure is no longer a cost center. It’s a value creation center. Companies are expanding and reallocating budgets because AI infrastructure drives growth, productivity and competitive advantage. Together with NVIDIA, we are giving enterprises the integrated platforms they need to move agentic AI from experimentation to results, on infrastructure they own and control,” said Michael Dell, chairman and chief executive officer of Dell Technologies.
New partner solutions and security
Reflection’s Beam, an open-weight model under the Apache 2.0 licence built for software engineering and long-running agent tasks, has been validated on the platform. Dell also added validated offerings from Dynatrace for AI observability, Mistral for a knowledge assistant and agentic platform, and Fogsphere for computer vision. SUSE Linux joins as an operating system option.
On security, Dell is offering a layered architecture that combines hardware-rooted infrastructure, NVIDIA’s OpenShell Reference Architecture for runtime governance and Red Hat AI Factory software for orchestration. A new Dell Managed Services for AI is meant to help firms govern access and control costs as they scale.
Agents arrive at the desk
With 81% of organisations planning, piloting or deploying AI-enabled PCs, according to an IDC paper cited by Dell, the company is widening its Deskside Agentic AI line with reference templates. These set out the hardware, models and software for uses such as agentic coding, and can scale from one developer to 1,000 or more systems.
A 64GB version of the Dell Pro Precision with GB10 targets individual developers and small teams, and units can be clustered in twos or fours. The Dell Pro Precision with GB300 is due on Windows, offering up to 748GB of memory and 20 petaflops of FP4 compute.
Servers for scale
Dell also unveiled data centre products: a containerised Modular Data Center built around NVIDIA’s DSX POD design; the PowerEdge M9882, which holds 256 NVIDIA Vera CPUs in one rack for running agent workloads; the air-cooled PowerEdge R8822 for standard racks; and the PowerEdge XE9810, a liquid-cooled rack using Groq 3 LPUs for high-volume inference. NVIDIA Aerial RAN Computer platforms on PowerEdge servers target telecom operators.
NVIDIA founder and CEO Jensen Huang said: “AI agents reason, use tools and learn—and they need a new kind of computer. Together, Dell and NVIDIA are building the computing foundation for the age of AI—from the deskside to the data center—so companies can build and run AI agents securely at industrial scale.”
Matt Kimball of Moor Insights & Strategy said customers are shifting focus from building infrastructure to producing measurable business results.
Availability
Beam, the Mistral offerings and the deskside templates are available now, along with the PowerEdge R8822 this month. Dynatrace and Fogsphere solutions arrive in the fourth quarter of 2026. SUSE Linux, Managed Services, the M9882 and the XE9810 follow in early 2027, as does the modular data centre.
