Michigan Tech’s Immersive Visualization Studio (IVS) provides researchers with the ability to visualize that data on an unprecedented scale. Located on the fifth floor of the EERC, it is a joint effort between Information Technology Services and researchers from the Paul and Susan Williams Center for Computer Systems Research.
“As high-performance computing plays an important role in sciences and engineering, advances in technology are helping researchers study increasingly complex phenomena and produce large amounts of data,” says Gowtham, HPC Research Scientist in Information Technology Services. “The use of single large screens for analyzing such data and visualizing results is neither commonly used nor economically feasible. Instead, a tiled display which uses several smaller, commodity displays to build a large screen – display or video wall – can help, as part of a high-performance visualization cluster (HPVC), in analyzing data and visualizing results.”
Built with Rocks Cluster Distribution, the 8 tile nodes (each with 4 CPU cores, 32 GB RAM and 2 NVIDIA GeForce GTX 680 GPUs) power the 24 46” Samsung LED monitors (each at 1920×1080 resolution and 240 Hz) for a stunning 50 megapixel overall resolution. The IVS also has 12 state-of-the-art motion tracking cameras.
Using this scalable tiled display approach allows the video wall to be more easily moved, the cost associated with repairing (or replacing) portions is lower, and there is no or minimal stoppage in research activity due to repairs.
Since reaching the production phase in Summer of 2012, countless hours have been invested, by ITS, to make it both a computational and visualization resource, and by researchers, to design award-winning scientiﬁc and engineering applications with real and measurable impact. The Immersive Visualization Studio oﬀers research and educational opportunities at every level and is expected to help cross-pollinate research ideas across various departments to help build a close-knit research community at Michigan Tech.
For a complete list of projects, visit http://ivs.research.mtu.edu/list-of-projects/.
Researchers interested in using this facility should contact Dr. Saeid Nooshabadi at firstname.lastname@example.org.