Brown University Department of Computer Science (BrownCS) alumnus Alex Tarvo PhD ’14 has just won the Association for Computing Machinery (ACM)’s Student Research Competition Award for his research into predicting the performance of multi-threaded programs. He’s the second Brown CS recipient of the honor (Dan Keefe won in 2005 at ACM’s SIGGRAPH conference), which was awarded at the International Conference on Software Engineering (ICSE).
“As we build computers with more and more cores,” Alex explains, “their behavior becomes increasingly complex and difficult to understand.” This challenge was the genesis of his award-winning work: Tarvo has designed models that predict hardware performance across a variety of configurations based on running a program in a single configuration.
“Alex does top-quality research,” says professor Steve Reiss, his thesis advisor. “He’s independent and handles all the details of his work with great integrity and trustworthiness. With these models, he solved a hard problem that I’d thought would be easy, and he did it the right way.”
Analysis and prediction are an area of computer science that has long been one of Tarvo’s key interests. “I’ve always been captivated by it,” he says. “I like to look at risk, to do statistical modeling. It’s just an area that I’m naturally curious about.”
It’s also a field whose applications are as wide-ranging as programming itself. “Alex’s research,” Steve Reiss explains, “can benefit anyone who wants to optimally configure web applications. Look at companies like Netflix, who constantly need efficiency from more and more complex data centers. These models can help system administrators foresee how performance will change with new hardware.”
Even as the award marks the close of a chapter in Tarvo’s academic life, the research it honors will undoubtedly continue to prove useful. “I feel great that he’s won this award,” says Reiss. “I’m confident that he’ll keep tackling similar problems in the next phase of his career.” Alex is eager to see both himself and others put his research to good use: “It’s satisfying to receive this after years of hard and sometimes stressful work! It’s great recognition, and I’ll be very excited if it helps make my research more available to people who can use it.”