Data processing group (PDV)


- Application fields
    - Computing infrastructure (networks, databases, virtualization, and cloud platforms) for scientific large-scale research
    - Data managment systems for energy storage research, neutrino physics, cosmic ray physics, meteorology, and fusion power 
    - DAQ and slow control systems for Neutrinophysics and dark matter searches
    - Distributed DAQ and control systems for high-bandwidth instrumentation in photon science
    - Real-time image processing applications in photon science and Material science
    - Highly parallel and reduced precision algorithms for high-speed synchrotron tomography
    - Smart scientific cameras for image-based data analysis

 - Technologies
    - High performance computing
	- Heterogeneous computing with CPUs, GPUs, FPGAs, and custom accelerators
	- Reduced-precision computations and hardware-aware techniques
	- Parallel programming with CUDA and OpenCL
	- Performance analysis and software optimization
    - High-bandwidth data acquisition, online data processing, and low-latency control
	- Linux kernel development, PCI drivers, Scatter-gather DMA engines, etc.
	- High-speed distributed storage with iSER, DRBD, Raid, and Linux AIO
	- Low-latency Infiniband and Ethernet networking using RDMA and ROCe technologies
	- GPU interconnects with NVIDIA GPUDirect and AMD DirectGMA
    - Data-intensive Cloud Platform
	- Data-intensive work-loads on OpenShift
	- GPU and desktop conteinerization
	- Scientific Workflow Engines
    - Experimental control, data management, and automation
	- Slow control platforms from EPICS and National Instruments 
	- Database technologies: relational (MSSQL, MySQL), document (MongoDB), time series (InfluxDB), image/matrices (TileDB)
	- Web-based data visualization with Spark, Jupyter, and WebGL
    - IT infrastructure for scientific experiments and cloud-based services
	- RHEL/CentOS and OpenSuSE platforms. Automation with Ansible
	- Distributed file-systems: GlusterFS, GlusterBlock, CePH, BeeGFS.
	- Virtualization with VMWare, oVirt, and KVM
	- CRI-O/Docker containers, OpenShift/Kubernetes orchestration, Helm package management

 - Projects (open & closed)
    - Open-source software
	- ADEI+Bora: 	Advanced Data Extraction Infrastrucure / data management for time-series data
	- WAVe:		Web-based Visualization of Volumes
        - ALPS: 	Advanced Linux PCI Services / FPGA-development support and drivers
	- UFO:  	Pipelined parallel image-processing framework
	- UCA:  	Universal camera abstraction library
	- Concert:	Control system for synchrotron tomography

    - Parallel algorithms and software
	- Fast parallel back-projection
	- Micro partical imaging velocemetry (uPIV)
	- GPU-based track trigger study for high energy physics
    
    - Current:
	- KaaS		Katrin as a Service
	- KATRIN SCS	Katrin slow control and data acquisition system
	- Weather Tower Data management for meteorologic instruments
	- ROOF          Online tomographic recontruction for object tracking
	
    - Complete:
	- UFO / STROBOS-CODE
	- uPIV
	- ASTOR / NOVA
	- KITcube 	Control System
	- TOSKA 	Control System
	

 - Technology transfer
    - none

 - Which infrastructure do we operate that is worth mentioning with the image on the website?
    - KaaS (Katrin as a Service): Cloud Platform for KATRIN experiment (production)

    - Parallel computing lab
	- Heterogeneous Computing Cluster with GPUs, FPGAs, and accelerators
	- High-speed Network Fabrics: 200 GBit/s Ethernet and Infiniband
	- Real-time Storage Systems with SSDs and magnetic storage

    - DAQ + slow control lab
	- FPGA-based DAQ systems
	- KATRIN test setup  – A test system for the KATRIN Detector data acquisition system
	- NI based slow control systems

    - Cloud computing lab (future)

 - Results:
    - GPU Computing
    	- Reviewing GPU architectures to build efficient back projection for parallel geometries
	- Balancing Load of GPU Subsystems to Accelerate Image Reconstruction in Parallel Beam Tomography
	- Investigation of the flow structure in thin polymer films using 3D µPTV enhanced by GPU
	- Evaluation of GPUs as a level-1 track trigger for the High-Luminosity LHC
	- An Open Source GPU Accelerated Framework for Flexible Algebraic Reconstruction at Synchrotron Light Sources
	- GPU-optimized Direct Fourier Method for On-line Tomography
	- A GPU-Based Architecture for Real-Time Data Assessment at Synchrotron Experiments

    - UFO
	- UFO – A Scalable Platform for High-Speed Synchrotron X-ray Imaging
        - Real-time image-content-based beamline control for smart 4D X-ray imaging
	- High-throughput data acquisition and processing for real-time X-ray imaging
	- UFO: A scalable GPU-based image processing framework for on-line monitoring
	- A PCIe DMA architecture for multi-gigabyte per second data transmission
	- Ultra-fast streaming camera platform for scientific applications

    - Data management and Control
	- WAVE: A 3D Online Previewing Framework for Big Data Archives
	- Focal-plane detector system for the KATRIN experiment
	- Advanced data extraction infrastructure: Web based system for management of time series data

    - Applications
	- Ground-based observations of thunderstorm-correlated fluxes of high-energy electrons, gamma rays, and neutrons (most cited ASEC paper)
	- Catalog of 2017 Thunderstorm Ground Enhancement (TGE) events observed on Aragats
	- Improved upper limit on the neutrino mass from a direct kinematic method by KATRIN
	- Parasitoid biology preserved in mineralized fossils
	- KITcube – a mobile observation platform for convection studies deployed during HyMeX
	- The NOVA project: maximizing beam time efficiency through synergistic analyses of SRμCT data.
	- Ultra-fast detector for wide range spectral measurements
	- Ultrafast linear array detector for real-time imaging