text.skipToContent text.skipToNavigation
|
NEW: NVIDIA DGX Spark – Local AI workstation meets server power
Artificial intelligence is developing rapidly. If the infrastructure cannot keep pace, delays are likely and many AI projects will ultimately fall short of their potential. Developers, data scientists, scientists and researchers in particular need local systems that are powerful, secure and flexible enough to efficiently develop and operate current and future AI applications – without having to go through the cloud or a data centre. With NVIDIA DGX Spark, the development of demanding AI workloads is now possible directly at the workplace for the first time. Typical use cases range from model development, fine-tuning and inference to accelerated data science workflows and edge AI applications in areas such as robotics and computer vision.
Highlights at a glance:
The DGX Spark enables local processing of up to 200 billion parameters – based on the NVIDIA GB10 Grace Blackwell superchip with 1 PFLOP AI performance and 128 GB of usable system memory. Two DGX Spark systems can be seamlessly connected via NVIDIA ConnectX-7. The pre-installed NVIDIA software environment allows for immediate prototyping and smooth scaling – whether in DGX systems, the DGX Cloud, or other infrastructures. Thanks to its compact desktop design, the DGX Spark fits flexibly into any development environment. At CANCOM, you can get the exclusive Founders Edition from NVIDIA as well as other models from leading manufacturers such as Dell, HP, Lenovo, and Asus. By adding NVIDIA AI Enterprise software, you get unlimited access to the complete software package for your AI workloads. You can also benefit from attractive pricing advantages for research institutions and start-ups as part of the NVIDIA Inception Partner Program.
More information
Start small, think big: AI server solutions with NVIDIA technology
Whether locally with DGX Spark or in the data center with compact AI servers, CANCOM enables you to scale seamlessly to productive operation.

AI SERVER FOR INITIAL PROJECTS

Server size S

This platform with two L4 GPUs from NVIDIA is suitable for smaller AI workloads and especially for inferencing. It is ideal for anyone who wants to test and further develop their own models—without any cloud dependency. It is space-saving and fits into an existing rack. It can also be used in a cluster at a later date.

To Product Overview

AI SERVER FOR PRODUCTIVE APPLICATIONS

Server size M

For those who want to start with specific use cases, the platform with two NVIDIA L40S GPUs offers an ideal basis. It covers small to medium-sized workloads such as fine-tuning, inference, or digital twin scenarios with NVIDIA Omniverse or the CANCOM Assistant. Thanks to its compact design, it fits into existing racks and can be expanded into a cluster if required. Those planning graphics- and visualization-intensive applications can rely on maximum performance with the new NVIDIA RTX 6000 Server Edition – perfect for 3D rendering, simulation, video production, and virtual reality.

To Product Overview

AI SERVER FOR SCALABLE WORKLOADS

Server size L

NVIDIA DGX systems or Dell PowerEdge servers with the latest NVIDIA Blackwell generation provide a powerful foundation for parallel, highly scaled high-end AI workloads. They are suitable for chatbots, computer vision, or LLM inferencing, for example, and create the foundation for an internal AI factory—a productive platform for developing, orchestrating, and deploying AI models.

Request consultation
CANCOM – Your strong partner for NVIDIA expertise, AI innovation, and data centers

CANCOM & NVIDIA

As a certified NVIDIA ELITE partner, CANCOM offers you much more than just state-of-the-art, high-performance hardware. On request, you can receive a tailor-made complete solution from a single source: we actively support you with AI enablement and training for your teams, provide comprehensive advice on the optimal solution, and are also available to assist you with comprehensive support during ongoing operations. We scale your solution as needed, from individual DGX systems to company-wide GPU clusters—as the foundation for your AI factory of tomorrow.

More information

We are your AI enabler in digital transformation

At CANCOM, we offer you a comprehensive end-to-end approach to the successful implementation of AI projects. Regardless of where you want to use AI—in the workplace, in production, or in customer service—we offer you a customized solution.

More information

Design your future-proof data center with us

Modern data center architectures are hybrid: in addition to the local data center, they comprise a mix of private and public cloud infrastructures and services. Our more than 200 data center experts work with you to develop an agile infrastructure and hybrid cloud strategy.

More information