Rutgers team provides computational support to research projects worldwide


unisupercomputers

Courtesy of Shantenu Jha | The RADICAL team uses a network of computers and supercomputers to help research groups calculate or simulate their work. Previous clients include members of the team that discovered the Higgs Boson.


Groundbreaking research requires a strong support network to provide clear and applicable results. That is where the Rutgers Advanced Distributed Cyberinfrastructure and Applications Laboratory comes in.

RADICAL, the colloquial nickname for the lab, has three subsections: the distributed computing team, computational science team and data-intensive team.  

Shantenu Jha, a professor in the School of Engineering and the leader of RADICAL, said the distributed computing team’s work involves bringing supercomputers and computers together. The team helped develop applications for those who work for the Atlas Project, the same people who discovered the Higgs boson, he said.

“That’s just an example of how the work this team does is really good to help a team plan the next 10 years of research and the experiments that they do," he said. 

The computational science team works on improving algorithms, methods and formulas that will help improve science, working to see how computers can be used more effectively in science, he said. 

The team works with people in pharmaceuticals, in particular looking at the conceptual ideas of protein structures and other critical structures, he said. 

The data-intensive team works with a variety of applications, helping people interested in the image-detection of cancer as well as neuro-imaging, he said. 

“(...) Many people say computational approaches are the third way of doing science. Theory, experiment and computation. So in some ways we’re facilitating that third way of doing science,” Jha said. 

Mark Santcroos, a graduate student and a member of RADICAL, said the team’s projects aim to bridge the gap technically and socially.

A critical part of the team’s work is accommodating a diverse set of users. Any given problem must be generic enough to be applicable to multiple cases.

“We think of the research challenges, we develop software to try to put our research ideas into practice, and then there’s a constant refinement once we apply them to applications,” Jha said. 

The applications the team codes must have real-world applications, as they solve real-world problems, said Matteo Turilli, senior research scientist at RADICAL.

"We have to create our tools to be able to do the measurements we need," he said.

The team finds these real-world problems in a variety of ways. They sometimes solve problems out of intellectual curiosity and publish the results. 

The team may also meet their potential users through conferences and suggest potential solutions to them, he said. 

“It’s very easy to start solving problems that are interesting but maybe not impactful on a meaningful time scale,” Jha said. 

When it comes to formulating the application, making it tangible can be difficult, said Ming Tai Ha, a graduate student with RADICAL. There are often a lot of small details that cannot be anticipated initially. 

“But once it works, it’s great,” said Tai Ha. 

When users can do things that they couldn’t go before, that’s a measure of success, Jha said. A successful project is simpler, more scalable, and innovative, Jha said. 

For example, the scale at which many of the protein simulators used now was inconceivable just a few years ago, he said.  

This past summer, Jha was invited to the White House to attend a workshop on science, technology and innovation, he said.

"They put us down right across the West Wing and we worked," he said. "We had to proffer opinions, and somebody took notes of what was being discussed. The notes are now being attached to the President’s One-Year Report on National Strategic Computing."  

Jha attributes his titles and success to his team, he said.

“Most faculty think they have a good team that supports them. They think they have the best team. I know I have the best team," he said. 'So, ask them if they think they work well enough to be called supercomputer gurus. I think they do. So really it’s the guru team.”


Faith Hoatson is a School of Arts and Sciences sophomore majoring in linguistics and French literature. She is a correspondent for The Daily Targum.


Faith Hoatson

Comments powered by Disqus

Please note All comments are eligible for publication in The Daily Targum.