Your Central Partner
for your new Data Center
work samples
Your standardized IT and Data Center checks
Comprehensive monitoring
Data Center Maintenance
Local AI systems: from applications to managed operations.
On-Premises & Edge
Dokumente, Prozesse & Assistenz
Access & data flows
Piloting, integration & support
By industry and application – from infrastructure to AI.
Data center certification is essential.
Blue DC A · Smart AI
From company knowledge to productive AI: Blue DC A brings GPU computing, power, cooling and controlled operations together. Up to six racks provide capacity for local AI, with liquid cooling directly at the chips.
AI-generated concept illustration. The configuration shown is not a completed reference project.
Blue DC A | Smart AI combines compact data centre infrastructure with the complete High Knowledge AI stack. We supply and integrate GPU hardware, local AI models, company knowledge and applications, right through to a digital employee for your team.
The building design envelope covers one to six IT racks with coordinated power and cooling. Your data, workflows and performance requirements determine the configuration you need.
Preliminary design assumptions, not verified performance figures for a completed Blue DC A installation: six racks at 120 kW each represent 720 kW of IT load. Applying an assumed factor of 1.25 for total facility power gives 900 kW; this factor is not a guaranteed PUE. Grid connection, reserve capacity, redundancy, cooling circuits and permissible rack loads require engineering verification. Smaller configurations start at a correspondingly lower capacity.
Technical reference for high-density AI systems: NVIDIA describes approximately 120 kW per rack and cold plates on CPUs and GPUs. This manufacturer example informs sizing; it is not a commitment to a particular hardware configuration.
High GPU density requires an end-to-end cooling path. We design water-based liquid cooling down to chip level: cold plates capture heat at the processors, and a separate cooling circuit carries it away. The design also accounts for residual air-cooled loads, leak detection and maintenance access.
Water temperatures, water quality and pressure limits depend on the selected systems. A heat reuse concept considers both the available temperature level and actual demand at the site.
Your digital employee searches approved documents, cites sources, drafts content and prepares emails or appointments. It works with your company’s data sources and access permissions. Your team reviews results and approves consequential actions.
Connect documents, systems and knowledge with traceable sources.
Integrate roles, permissions and approvals into the application.
Use search and assistance productively; have critical results reviewed by qualified people.
We support the entire journey: selecting use cases, connecting data sources, setting up hardware and models, testing permissions and training your team. This includes coordinated connections to email, calendar, CRM and other applications.
Within the agreed support scope, we provide monitoring, tested updates, model maintenance and application development. Backup, recovery, responsibilities and acceptance criteria are agreed together.
A low PUE alone does not demonstrate climate neutrality. Assessing a climate-neutral data centre requires a transparent greenhouse gas inventory with a defined reporting period and boundaries. Electricity procurement, refrigerants and the lifecycle stages included must be stated, alongside documented emission avoidance and reduction measures. Blue DC A does not itself establish that a facility or its operation is climate-neutral. For Blue DC A, we assess GPU utilisation, local AI models, power supply, liquid cooling and potential heat reuse.
Explore data centre planning →
