Research infrastructure

Our research infrastructure supports data-intensive simulations, the training of AI models, the development of control algorithms, and hardware-related prototyping and system testing. It comprises modern server and HPC resources, a testbed for cyber-biological systems under real-world environmental conditions, and a fully equipped electronics and prototyping laboratory for the development of embedded systems.

Our lab provides modern infrastructure for the development and analysis of embedded systems, digital hardware, and small-scale power electronics. It includes fully equipped measurement workstations, debugging tools, and equipment for rapid hardware prototyping.

Digital Measurement Equipment

  • Digital oscilloscope and digital multimeters
  • Differential probe for power electronics
  • Function and signal generator
  • Logic analyzer
  • Various debuggers and adapters for microcontroller development

Prototyping Equipment

  • SMD soldering station and hot-air rework station
  • ESD workstation including fume extraction
  • Bambu Lab X1 Carbon for rapid 3D printing of prototypes
 (opens enlarged image)ASN laboratory workstation for electronics development

Our computing infrastructure supports data-intensive simulations, numerical optimization, and the training and testing of neural networks. It includes several GPU and CPU servers across different performance classes as well as an energy-efficient Apple-Silicon mini-cluster. Together, these resources enable both HPC workloads and distributed experiments as well as virtual development environments.

Apple-Silicon Mini-Cluster

  • Specifications:
    • 3 × Mac Mini M2 (8 GB RAM)
    • 2 × Mac Mini M2 Pro (16 GB RAM)
  • Use cases: Parallel simulations, AI prototyping, and energy-efficient development workloads
  • Funding: Appointment funds of the University of Kassel

Server „hippo“ (High-end training & HPC)

CPU2 × AMD EPYC 9755 (zusammen 256 Cores / 512 Threads)
Memory1.5 TB DDR5 RAM
Storage5 × 7.68 TB SSD
GPUs2 × NVIDIA H200 NVL GPUs (je 141 GB, NVLink)
Use casesLarge AI models, neural network training, HPC simulations, large optimization problems
FundingEquipment Fund, University of Kassel
UsageShared with other research groups at the University of Kassel

Server „rhino“ (GPU-accelerated AI workloads)

CPUAMD EPYC 9555 (64 Cores / 128 Threads)
Memory384 GB DDR5 RAM
Storage2 × 7.68 TB SSD
GPUs2 × NVIDIA A30 GPUs (je 24 GB)
Use casesTraining and inference of AI models, simulations, solving optimization problems
FundingEquipment Fund, University of Kassel
UsageShared with other research groups at the University of Kassel

Server „elephant“ (Virtualization & software testing)

CPUAMD EPYC 9374F (32 Cores / 64 Threads)
Memory192 GB DDR5 RAM
Storage4 × 3.84 TB SSD
Use casesVirtual machines, continuous integration, simulations
FundingAppointment funds of the University of Kassel
(opens enlarged image)

Our testbed for cyber-biological systems consists of sensor- and actuator-equipped raised garden beds as well as a dedicated weather station. It enables us to investigate and validate system identification and control concepts, such as automatic, resource-efficient irrigation, under real-world environmental conditions.

Features

  • Four raised bed environments, each equipped with
    • 2 soil moisture sensors (ThetaProbe ML2x)
    • 2 temperature sensors
    • 2 pumps for local irrigation
  • Weather station (based on the Raspberry Pi Weather Hat) with sensors for
    • Wind speed
    • Wind direction
    • Temperature
    • Humidity
    • Solar irradiance
    • Precipitation

We thank the Central Teaching Fund of the University of Kassel for supporting the establishment of this testbed. We would also like to thank the many students who made significant contributions to bringing the environment to life.

(opens enlarged image)