German computer scientist Torsten Hoefler, a 34-year-old devotee of math and running, has won one of four 2015 Latsis Prizes for his research on high-performance computing.
Based at the Swiss federal technology institute ETH Zurich, Hoefler is internationally regarded as a young scientific leader in the field of high-performance computingexternal link by combining theory and application at his Scalable Parallel Computing Laboratory.
The computer scientist, who is an assistant professor at ETH Zurichexternal link and has long been fascinated by numbers, previously taught and conducted research in the United States where he worked on developing one of the world’s most efficient supercomputers. He began his academic career studying for a master’s degree at Germany’s Chemnitz University of Technology.
Hoefler won one of four university Latsis Prizes awarded each year by Swiss institutions, each worth CHF25,000 ($24,500). The National Latsis Prize, one of Switzerland's most prestigious scientific awards, is due to be announced on November 24. The prize carries an award of CHF100,000 and honours the outstanding scientific achievements of a research scientist under age 40 working in Switzerland.
The Latsis Foundation’s website cited Hoefler for outstanding “contributions to performance modelling, simulation, and optimization of large-scale parallel applications; topologies, routing, and host interfaces of large-scale networks; and advanced parallel programming techniques and runtime environments”.
On its website, ETH quoted Hoefler as constantly trying to find new ways to use numbers to improve his life, even going as far as creating a performance model for himself. It’s something he started doing as a child, when he would memorize as many car registration numbers as possible or count the distance to school in steps.
“I’ve definitely taken a bit of a mathematical view of life, but then my job is derived from my life,” he was quoted as sayingexternal link.
Hoefler has tried to unite theory and practice whenever possible, by developing mathematical models that can be translated into software for running some of the nation’s supercomputers. He and his team have focused on developing a so-called heterogeneous compiler that can translate and optimize applications for computer architectures.
He also enjoys running, which he considers to be another passion. He calls it useful not only for maintaining a healthy body and mind, but also for discussing problems with students who join him on runs.