Computer and projection infrastructure for visualization in scientific projects, such as the analysis of measurement data, illustration of computer simulations in the field of Computational Fluid Dynamics or in computer-aided product development at Campus North.
ForHLR Visualization Lab at Campus North
In order to meet the ever-increasing demands on visualization and at the same time the rapidly growing amount of data in scientific projects such as the analysis of measurement data, illustration of computer simulations in the field of Computational Fluid Dynamics or in computer-aided product development, the SCC has acquired both a visualization cluster and workstations for the visualization of the results on the projection of the visualization laboratory at Campus North with the second phase of the ForHLR research high performance computer. Thus, visualization laboratories are available on both the South and North campuses.
The new visualization lab in the North Campus is suitable for interactive work in small teams, for demonstrating interactive work in front of larger groups, as well as in lecture scenarios with a seated auditorium. The rear projection screen has a size of 6700mm x 2300mm with a resolution of more than 13 megapixels.
The three visualization workstations are each equipped with at least
- 2x Intel Xeon E5-2630v3 processors, 2.4GHz, each with 8 cores
- 1-2 NVIDIA K6000 graphics cards
- RAM: 512 GB RAM
- 4x1 TB SSD disks each
With the two 4K projectors, the image is projected side by side with a so-called blending zone from behind onto the projection screen and thus brought to its full width. The following scenarios are possible:
- 3D Fullscreen
3D working environment, maximum area
- 2D "Cinerama" mode
A recording is shown on the maximum possible area.
- Multi-Display (2D / 3D mixed)
At least 2 recordings are visible on the projection screen,
- Lecture scenario
Picture bottom edge 90 cm, maximum possible area
The visualization software products Covise, Ensight, ParaView and Vapor are available under Redhat Enterprise Linux and Windows. Further products can be provided according to requirements. For the remote visualization of result data from the visualization cluster on the projection or on local workstations we recommend the use of the script start_vnc_desktop. You can find a description of this script at https://wiki.scc.kit.edu/hpc/index.php/Remote_Visualization or with "start_vnc_desktop --help".