CCV hosts high-performance computing resources

Before researchers can answer questions about the impact of climate change on marine life and ecosystems, they have to collect data — reams of facts and statistics that must be stored and studied, assessed and analyzed.

But, as anyone with a household of Wi-Fi dependent gadgets might be able to relate, harnessing technological output, whether sending, streaming or downloading, demands capacity and efficiency.

Alone, Rhode Island scientists and their respective institutions can’t afford what they need. But together, through the EPSCoR effort to share resources, they can tap into the high performance computing and research data services of the Center for Computation and Visualization (CCV) at Brown University without having to leave their laptop.

“One aspect, certainly, is sheer speed. But now, there also is the qualitative aspect. It’s not just that we couldn’t do it before and now we can do it faster. Now, people can ask bigger questions, questions that weren’t practical to ask.”

Housed at Brown’s Providence campus, the CCV and its vast capabilities can be accessed by RI NSF EPSCoR researchers at any one of the nine partner institutions across the Ocean State. The high-performance computing resources equip the research community to undertake complex numerical simulations, modeling, and data analysis.

Part of Brown’s Computing and Information Services (CIS), the CCV fosters and manages high performance computing, highly reliable research data storage, visualization resources, physical and virtual server hosting, high-performance backup and archival services, and scientific support to empower computational research, scholarship and creativity, according to CCV Executive Director Gurcharan Khanna.

“We have large scale computers that are faster and can handle more data than what may be available on their campus, whether they’re processing numbers or creating visualizations,” explains Khanna. “We work alongside researchers to optimize their workflow and get their research done efficiently.”

RI NSF EPSCoR collaborates with the CCV to promote the use of cyberinfrastructure, computational science, and bioinformatics in studying the responses of marine life to climate variability. The initiative has marshaled the forces of the research community and led to integration of resources.

EPSCoR researchers have access to a dedicated portion of the computer cluster for high performance computing and data services for data analysis. If a research project aligns with the EPSCoR questions, then every researcher and his or her group can have accounts set up to access the system. With virtual server hosting, high-performance backup and archival services, scientific support and pre-installed software tools, the facility provides an invaluable platform for scientific research.

High throughput analysis holds the key to solving many critical research questions, and recent advances have decreased the cost to procure data. Yet, with more data, the need increases for data management, analysis and archival storage.

The CCV allows scientists to log in, whether from a terminal in their office or on the go with a mobile device, and connect to powerful technology. Equally important, the CCV provides individual support throughout the course of the workflow.

“Whatever it takes to help get the science done,” Khanna says. “The CCV is really about people first — the researchers and our staff, who are often themselves research scientists, helping researchers use the technology effectively.”

However, capacity alone cannot meet the demands of scientific pursuit to answer the complex and compelling questions faced by humankind. Researchers also must be able to transmit the huge amounts of data collected, which is why RI NSF EPSCoR in the beginning grant years made significant investments to improve the state’s cyberinfrastructure and develop resilient, high-bandwidth connectivity between research and academic institutions.

Those early partnerships and collaborations between EPSCoR and other Rhode Island organizations means researchers today throughout the state can upload big batches of data for processing and storage at the CCV.

To gain a sense of what this means, think about the latest and greatest in personal computing — a typical laptop may have a storage capacity of 250 gigabytes. The CCV currently has the capacity of about 2,000 such laptops or 500 terabytes, which equals a half of a petabyte, and is considering requests for one petabyte, which equals 1,000 terabytes or 1 million gigabytes.

Once the data arrives at the CCV, a computer program there can sort, manipulate and visualize the data, and all at a high speed.

“Even if you’re dealing with 100 gigabytes of data,” explains Khanna, “you want to make sure you’re not moving it at Wi-Fi speed.”

This is a critical point because data, once collected, is in a continual state of movement in the workflow. Data gets gathered, stored, modified, moved, modified again, and analyzed. Results might be brought back to an individual’s computer; results might be disseminated and then sent back again.

As Khanna notes, with all this movement, it doesn’t matter how fast the long haul is if the point of use is slow. To make all parts compatible and useful, from collection to processing and, ultimately, storage, the system needs both the high speed of fiber optics plus the equipment capacity along the way to handle the delivery. If researchers need additional capacity at the CCV, the facility can add on storage.

In all of the talk of terabytes and moving ginormous quantities of data at high speeds, Khanna finds greater meaning in what this system allows.

First, from an economic standpoint, there exists significant value in a shared system. What the CCV offers is bigger and more capable than what any single person or institution can afford alone.

There also is the factor of speed and what that allows researchers to accomplish. What used to take an hour, now may consume only a single second; or, what previously took years, today may take minutes. So, imagine the possibilities.

Yet, there is more to consider than merely speed and function, as exciting as the potential promises.

Khanna reasons, if the process takes a long time, researchers may not be compelled to ask the questions. So, the capacity alone prompts scientific pursuit.

“One aspect, certainly, is sheer speed,” says Khanna. “But now, there also is the qualitative aspect. It’s not just that we couldn’t do it before and now we can do it faster.

“Now, people can ask bigger questions, questions that weren’t practical to ask.”

And therein lie the beauty and the wonder, and the possibility. Not only do we have the technology to support our quests, but also where else might we now be able to go?

By Amy Dunkle | from the Spring 2015 issue of The Current

Photos courtesy of the CCV