Imagine strolling along a quiet hallway on the second floor of Pasteur Hall and noticing a group of Bellarmine physics students having a lively discussion with their mentor, Dr. Akhtar Mahmood, all while staring and pointing at a large display wall in what looks like a NASA control room.
That room is actually Bellarmine University’s new state-of-the-art Advanced Visualization and Computational Lab, which was funded by a $209,347 grant received by Dr. Mahmood in 2012 from the National Science Foundation. The grant, funded under the Obama administration’s R&D initiative, is aimed at advancing the scientific and technological means of managing, analyzing, visualizing and extracting useful information from Big Data.
Tiled walls that display large quantities of information at a single glance have traditionally been the domain of Mission Control, TV news channels, big-time sports stadiums and big-budget corporations. But in recent years, these display walls have also been installed at a few research universities and national labs.
Stephen Brown, an adjunct faculty member in the Physics Department, physics majors Amber Byrum, Bryan Fagan and Tyler Martinez, Derick Hill from Bellarmine’s Technology Services, and Mediatech helped Dr. Mahmood set up the lab.
“Scientific inquiry in the 21st century will depend in large part on data exploration,” Dr. Mahmood said. “Finding new information and patterns in Big Data is the new imperative. Large-scale data-driven applications are on the rise, and so is the need to extract specific information from terabyte- and petabyte-scale data sets,” or Big Data. In the case of high energy physics experiments, for example, extracting rare “discovery” signals by separating background events from large data sets is a very challenging task. And along with the rapid growth of the size of these data sets, the resolution of the data from the large scientific instruments is also increasing.
Understanding these multi-dimensional, time-varying data sets, requires advanced visualization and analytics tools.
Bellarmine’s Advanced Visualization and Computational Laboratory houses a high-resolution (16 mega-pixel) multi-screened tile wall 16 feet wide and 5 feet tall, called Hiperwall (Highly Interactive Parallelized Display Wall). The Hiperwall is connected to nine high-end workstations equipped with advanced multi-core graphics cards that allow students to conduct large-scale data analysis and visualization studies of data coming from the ATLAS high energy experiment at CERN to Bellarmine via the national Open Science Grid. Dr. Mahmood and his students are also using the visualization display wall as a Tier3 Grid Command Center for Bellarmine University’s supercomputer and to monitor all the ATLAS grid jobs that are being processed worldwide.
The key feature of the display wall is the bezel size of the display tiles. Arranged in a 4×2 array, each tile is a high-end, commercial-brand 55” Samsung backlit-LED HDTV that has a very narrow bezel size of about one-tenth of an inch. “That’s the narrowest bezel size you can get these days,” Dr. Mahmood said.
Each of the Hiperwall display node PCs is responsible for displaying and rendering a portion of the overall image that is being displayed on the display wall. Students can control the Hiperwall visualization system remotely via a tablet, and the technology allows researchers in geographically diverse locations to collaborate on large scientific experiments that generate Big Data.
Bellarmine was the first undergraduate institution in the United States, and the only institution in Kentucky, to implement the Hiperwall technology to analyze data from the ATLAS high energy physics experiment.
Users can display a single image (up to 1 GB in size) across the entire display or many different images simultaneously. Each image can be independently sized, positioned and rotated. A wide variety of high-resolution 2D and 3D images, animations, movies and time-varying data can be displayed in real time all at once on multiple display tiles, according to the needs of the users. For video, animations and streaming movies across the network, the Hiperwall visualization system’s software can provide feeds of up to 60 frames per second to the display wall.
In industry, such visualization tools can provide a competitive edge by transforming business and engineering practices. Visualization of data events often provides the only means of communication between the complex and abstract world of particle physics and the general public, making visualization a valuable educational and public relations tool.
“The Hiperwall system has created a state-of-the-art visualization environment by stimulating and engaging our students in inquiry-based visual learning, and for conducting interactive and exploratory data analyses activities,” Dr. Mahmood said.