text.skipToContent text.skipToNavigation
CLOUD MARKETPLACE | SHOP |
|
  2. Dell Pro Max GB10

With the Dell Pro Max featuring NVIDIA GB10, you can bring AI development and inference directly to the workplace – combined with CANCOM services for implementing a local AI assistant based on NVIDIA NemoClaw and Nemotron 3 Super (120B).

Powered by NVIDIA Grace Blackwell, the system can deliver up to 1 petaFLOP of FP4 AI performance in a compact, energy-efficient form factor. With 128 GB of coherent unified memory, companies can run larger models and autonomous AI agents locally, scale configurations up to four times, and leverage the full NVIDIA AI software stack as well as its entire ecosystem.

The energy-efficient design is ideal for continuously running agents. Dell Technologies and NVIDIA are also jointly developing an air-gapped solution for government organizations that enables autonomous AI agents – based on classified and sensitive data – to operate in physically isolated environments without external network connections.

This bundle includes:

HARDWARE

Dell Pro Max with NVIDIA GB10

  • Local AI power directly at the workplace – for prototyping, fine-tuning, and local inference of large models (up to ~200B parameters)
  • 1 PFLOP FP4, 128 GB LPDDR5x unified memory, up to 4 TB NVMe
  • NVIDIA DGX OS (Ubuntu-based), flexible connectivity options including 10GbE/Wi-Fi 7 and ConnectX-7 SmartNIC

SERVICE

CANCOM Introduction “Local AI Assistant” with NVIDIA NemoClaw

  • Installation/basic configuration of the NemoClaw stack, including OpenShell Runtime, sandbox/policy foundations, and local inference setup
  • Focus on control, privacy, and secure execution: “built-in safeguards” and configurable security policies
  • Optional extensions available: use case workshop, operational handover, support/hardening

AI MODEL

Nemotron 3 Super (120B)

  • 120B parameters (12B active), optimized for agentic workflows (reasoning, tool use, planning) – freely available under the NVIDIA Nemotron Open Model License

Which areas benefit most from this bundle?

✓ Engineering Copilot (local): Code assist, refactoring ideas, documentation drafts – without data leakage
✓ IT Automation & Runbooks: Agent-based task processing (planning/tool use) based on an openly available AI model with a strong reasoning focus.
✓ Document workflows: Summarizing, structuring, drafting – locally, controlled, and with integrated security mechanisms.
✓ Prototyping/Fine-tuning at the workplace: Short development cycles and inference without detours via external services or infrastructure.
Be productive faster – without giving away your data
  • Data sovereignty: Workflows and data remain within your environment – especially important for internal and sensitive use cases.
  • Low latency & full control: Local inference without routing through external services – ideal for iterative development and fast results.
  • Scalability: From local prototyping at the workstation to later scaling in infrastructure environments – with a clean transition to suitable target architectures.
I consent to the processing of my personal data. For more information about data processing, please click here.
Are you interested? Please contact us directly!
We are happy to advise you on the Dell Pro Max GB10 and show you how to securely and efficiently deploy local AI in your company.