High-powered computing superhero
“If there’s a need for computing power to do research, I can probably help.”
Julie Faure-Lacroix, a bright-eyed young woman from Laval University, is talking about her job at Calcul Québec, a regional partner of Compute Canada. Although she never says it herself, it’s clear she’s a bit of an academic’s superhero, coming to the computing rescue just when researchers think their hopes for a project have been dashed.
As a scientific liaison agent, her job is to reach out to academic researchers to show them how high-powered computing can help them in their research. The mandate of Calcul Québec and Compute Canada is to lead the acceleration of research and innovation by deploying state-of-the-art advanced research computing (ARC) systems, storage and software solutions.
Faure-Lacroix has a master’s of biology from Laval University, but she started working for Calcul Québec while working on her PhD. Her PhD research involves studying bats whose populations have been severely curtailed by white-nose syndrome, with a goal of contributing to the establishment of a management system for a viable population.
But by day, she helps students and academics do their work more efficiently and effectively with the use of computers. The fact that she has a science background has helped her to reach out more effectively to biologists and other scientists who might not understand the benefits of high-powered computing, but who understand it when she explains it because she knows exactly what they need.
“In biochemistry and genomics, people have been using high-powered computing for years,” she says. “But those who study mice and parrots and birds haven’t necessarily. They’re in their labs, hidden away. They don’t know they can use supercomputers for their work and Calcul Québec had trouble reaching out to them because they didn’t know them or how to speak their scientific language. I’m essentially a translator and it works very well.”
Asked to give an example of her work that she would use if someone at a cocktail party asked her what she did, she referenced a project she’s done with Quebec’s ministry of forest, wildlife and parks.
As part of a massive project studying climate change, they are building maps and trying to predict what future forests will look like. To do that, they look at other places in North America, places that currently experience the temperatures that will eventually arrive in Canada.
But to look at all of North America translates to a ton of data, so Faure-Lacroix helped them reduce their computing times considerably. In 2009, when they ran their models on their existing systems, it took them six months to process the data. Today, they can do it in fewer than 48 hours.
“That’s just by changing the code, using better machines and because of the benefits of parallel computing,” she says with a smile.
Another example: A professor at the University of Montreal contacted her about a student who wanted to run models that were taking months and months to run. The professor was worried his student would run out of time to complete his master’s degree. Faure-Lacroix came to the rescue and not only helped him run his models more efficiently, she also showed him how to transform his code into parallel code, allowing him to run much more complex models, making his research more in-depth and, ultimately, more meaningful.
“We often find people give up because their computers crash or it takes forever, so they reduce the scope of their research or aggregate the data,” she says. “This is what we want to avoid. The mistake is using a laptop to do something the scope of which only your mind can imagine. That’s when I’m useful.”
Faure-Lacroix says she has two distinct advantages in her job. First, being a woman, she thinks, makes her less intimidating, and second, being a scientist means she speaks the language of those she seeks to help.
“I love the helping part of my job,” she says. “It’s very rewarding. When people at conferences are recommending you to other researchers, you know they really appreciate what you’ve done.”
And what’s next for her? She hopes to complete her PhD and one day become a professor who uses the services of Calcul Québec and Compute Canada. In the meantime, she’ll keep helping like-minded scientists.