Cloud&Heat Atlas is a comprehensive on-premises cloud infrastructure for businesses requiring a high-performance, reliable and sovereign infrastructure for AI and data-intensive applications. The solution is provided as a ready-to-use appliance, including hardware specifically designed for the respective use case and featuring state-of-the-art GPU accelerators. The hardware is typically supplied with liquid cooling for high energy efficiency.

By combining high-performance hardware, cloud software and operational services, businesses gain a ready-to-use infrastructure that significantly reduces the effort involved in planning, integration and operation.

Cloud&Heat Atlas supports the entire AI value chain – from data preparation and model development, through training, to the production deployment of AI applications. The infrastructure can be used both for internal AI workloads and for the delivery of applications and services as Software-as-a-Service (SaaS).

Cloud&Heat Atlas is particularly well suited to production-grade inference and real-time applications where high availability and reliable operation are crucial, and where there are stringent requirements regarding data protection and performance.

For resource-intensive AI workloads, the platform also offers the option of particularly energy-efficient operation through the use of direct liquid cooling.

The software stack is based entirely on open-source components. The technical foundation of Cloud&Heat Atlas is OpenStack, which serves as the cloud management platform; specifically, it is based on the distribution Cloud&Heat Patron. To ensure automated and scalable operation, OpenStack is deployed on Kubernetes using Yaook.

Optionally, a Kubernetes platform based on the distribution Cloud&Heat Compass run on the OpenStack infrastructure. This enables containerised applications and modern cloud-native workloads to be deployed and managed flexibly.

Cloud&Heat Atlas is based on established and widely used cloud standards in the areas of interfaces, security, interoperability and portability. On request, the OpenStack infrastructure can also be set up in accordance with the Sovereign Cloud Stack (SCS) guidelines. This creates additional transparency, promotes interoperability between cloud environments and helps organisations meet requirements for digital sovereignty and long-term investment security.

Yes. Cloud&Heat Atlas is fully installed, configured and handed over ready for use by Cloud&Heat. As part of the commissioning process, we offer training and knowledge transfer on request, so that your team can administer and operate the infrastructure independently.

Once the handover is complete, operations can be carried out entirely by your own staff, provided you have the necessary staff resources and expertise. Our on-demand support is available at any time to answer any questions, carry out maintenance tasks or provide technical assistance.

Alternatively, we can take over all or part of the operation for you. Our service options range from extended support services during business hours to a comprehensive 24/7 operation and monitoring service.

Cloud&Heat Atlas is designed as an integrated, end-to-end solution. The hardware is specified in consultation with you, based on your requirements, and optimally tailored to the specific application. This ensures the best possible performance, energy efficiency, scalability and operational reliability.

If you already have hardware in place, we would be happy to assess its technical suitability. Depending on the resources available, we can provide a comparable cloud infrastructure based on our bespoke on-premises solutions. This is subject to the existing hardware meeting the requirements of the desired platform architecture.