Dell Technologies introduces Dell Deskside Agentic AI, a solution that enables organizations to build and run autonomous AI agents locally on high-performance workstations. Data remains within the organization, and costs are predictable. The solution is also scalable to the data center via a common framework.
The solution is part of the Dell AI Factory with NVIDIA and combines Dell workstations with the NVIDIA NemoClaw software stack. NVIDIA OpenShell is now available across the entire AI Factory, from workstations to PowerEdge XE servers. You can build, test, and refine the AI agents within a sandbox environment.
Three hardware configurations
Dell offers three workstation configurations, with each model becoming progressively more powerful and capable of handling larger, more powerful workloads.
The Dell Pro Max with GB10 is the compact entry-level option for individual developers and small-scale agent prototyping. This entry-level model supports models with up to 30 billion parameters. Next is the Dell Pro Precision 9, an enterprise workstation tower with Intel Xeon 600 processors and up to five NVIDIA RTX PRO Blackwell GPUs. This model is designed for models with 30 – 500 billion parameters. Finally, there is the Dell Pro Max with GB300, which features an NVIDIA Grace Blackwell Ultra Superchip. This is the most powerful variant and supports models with 120 billion to 1 trillion parameters.
NemoClaw and OpenShell
The NVIDIA NemoClaw software stack serves as the solution’s software backbone. It is an open-source framework for managing always-on AI agents built on OpenClaw. The stack combines Nemotron models for reasoning and coding, OpenShell for policy-based security controls during runtime, and an Agent Toolkit for the orchestration of long-running agents.
From the desktop to the data center
Dell is also adding support for NVIDIA AI-Q 2.0, a blueprint for deploying multi-agent workflows for research, decision support, and complex tasks. The Dell-NVIDIA AI-Q 2.0 Reference Architecture on the Dell AI Data Platform is specifically designed for regulated industries, such as financial services, the public sector, and manufacturing. With this support, agents designed and built on a workstation can be deployed on data center infrastructure that is much more powerful and has access to much more data.
Cost Claim
Dell claims that organizations can save up to 87 percent on cloud costs over two years. That calculation is based on a report commissioned by Dell from Signal 65 and Futurum Group. The analysis assumes a combination of Dell Pro Max workstations and PowerEdge servers for general, sales, and software development workloads over a five-day workweek, including estimated cloud discounts and infrastructure costs. A remarkably high figure that is difficult to verify; however, it is true that organizations with an on-premises infrastructure are less likely to face unexpected costs. The cloud bill cannot suddenly skyrocket, something that some boardrooms do fear.
Availability
Dell Deskside Agentic AI is available immediately. NVIDIA OpenShell and NVIDIA AI-Q 2.0 for the Dell AI Factory are now available, as is the Dell-NVIDIA AI-Q 2.0 Reference Architecture.