University of Stuttgart tests Databox connection to the NADIKI Registrar in the lab

University of Stuttgart tests Databox connection to the NADIKI Registrar in the lab

The IER and IAAS of the University of Stuttgart have developed a data box that collects energy and environmental metrics from the data center infrastructure and transmits them to the NADIKI Registrar. The initial functionality test on a laboratory scale was successful.

The Institute for Energy Economics and the Rational Use of Energy (IER) and the Institute for Architecture of Application Systems (IAAS) at the University of Stuttgart jointly developed a Databox and installed it in the IER server room. The Databox collects metrics from the physical infrastructure — power consumption, temperature, and humidity — and transmits them to the NADIKI Registrar. The first functional test on a laboratory scale was successful.

The Databox addresses a central issue of the Observer Architecture: How can measurement data from existing data center infrastructure be converted into a standardized format for the Registrar? Robin Pesl and Dinesh Vemula from IAAS conceptualized and built the Databox architecture. The IER provided the laboratory with Lenovo Edge servers and full infrastructure (cooling, UPS) as a test environment.

Databox-Installation im Server-Rack des IER-LaborsTestumgebung im Labor des IER an der Universität Stuttgart
Research
NADIKI - Sustainability Indicators for AI

Within the NADIKI Project, we are integrating Environmental Impact Indicators from the Life Cycle Assessment Methodology into Kubernetes. This empowers workloads, like AI training workflows and inference, to assess their operational impact and allows detailed reporting. Using this data, we evaluate the feasibility of an environmental product declaration for an AI model.

Research
NADIKI - Sustainability Indicators for AI

Within the NADIKI Project, we are integrating Environmental Impact Indicators from the Life Cycle Assessment Methodology into Kubernetes. This empowers workloads, like AI training workflows and inference, to assess their operational impact and allows detailed reporting. Using this data, we evaluate the feasibility of an environmental product declaration for an AI model.