The data centre enables fast and efficient work with data within the technologies of an integrated data centre – Apache Hadoop, iRODS and others. Thus, it allows the processing of large data sets (Big Data) from various sources – medical data, data collected from transport and traffic flows, industrial production, smart cities or banking. The data processed in this way can then be further analysed and visualized as required within the specified research task.
Support for research and development in the field of parallel optimization, processing and consolidation of data.
Total calculated power: min. 160 TFLOP consists of min. 80 TFLOPS CPU + min. 80 TFLOPS when using the accelerator. They also include:
- Control nodes [2pcs] -specification and architecture corresponds to computing nodes with 128 GB RAM.
- Login nodes [4pcs] -specification and architecture corresponds to computing nodes with 768 GB of RAM.
The system includes parallel and shared storage with 4 PB of total usable capacity with support for parallel and shared file system. Such a solution contains several software packages that allow you to compile and run your applications.
- Compilers – allow the compilation of user software that uses CPUs or accelerators.
- Runtime libraries – MPI libraries enabling linking and software execution
- requiring MPI
- Big Data processing – placing software packages as
- Hadoop, Map-Reduce and Chandeliers based on current on-demand requirements for individual compute nodes to enable the execution of Map-Reduce-based applications.
- Remote Visualization – allows users to visualize data and results of scientific calculations using computational tools available at these nodes.
- Visual analysis – is statistical and Visual software that enables data analysis for researchers. This software is equipped with standard data connectors for various data sources.
- Extensive graphical development platform – integrated development environment and libraries
- Support platform of 2D and 3D modelling – allows a connection between computing nodes, workstations and data storage.