Indiana University alumnus Torsten Hoefler (Ph.D. 2008) was recognized for his outstanding contributions in the application of high performance computers using innovative approaches at the International Conference for High Performance Computing, Networking, Storage, and Analysis (SC22) in Dallas, Texas, US. The 2022 IEEE Computer Society’s Sidney Fernbach Award honors Hoefler’s contributions in the application-aware design of high performance computing (HPC) algorithms, systems, and architectures, and his transformative impact on scientific computing and industry.
The award was established in 1992 in memory of Sidney Fernbach, a pioneer in the development and application of high performance computers for the solution of large computational problems. Hoefler is the youngest recipient to date.
Dr. Torsten Hoefler, IU Luddy School of Informatics, Computing, and Engineering alum
I love working with young people, I can see their energy and would like to give them more energy to move forward, especially when they’re in the early career stage. In that stage, they’re still open to completely crazy ideas.
Now a professor of computer science at ETH Zurich in Switzerland, Hoefler began his work at IU in 2005 under the guidance of Professor Andrew Lumsdaine in the Luddy School of Informatics, Computing, and Engineering. His work on the Message Passing Interface (MPI) and various communication networks in Lumsdaine’s group at IU laid the foundation for his future achievements. His doctoral dissertation topic included various works on implementation and specification of nonblocking collective operations, which later played an important role in deep learning and are used by hundreds of thousands of researchers and practitioners today.
Since then, Hoefler has been instrumental in developing techniques to improve the efficiency of high performance computation and large closed data systems. Many of his ideas have helped to shape the core components for constructing, running, and programming supercomputers, like the new ALPS system, which is due to come online in spring 2023 at the Swiss National Supercomputing Centre.
Hoefler’s work primarily focuses on improving the performance of highly scalable parallel systems and developing numerous applications in the areas of weather and climate simulation, as well as machine learning.
Hoefler is also known as an enthusiastic professor, inspiring future generations with his excitement for trying new things. Hoefler serves as director of the Scalable Parallel Computing Laboratory at ETH Zurich, performing research in all areas of scalable computing, including high performance networks and protocols, middleware, operating system and runtime systems, parallel programming languages, and support.
“I love working with young people,” he said. “I can see their energy and would like to give them more energy to move forward, especially when they’re in the early career stage. In that stage, they’re still open to completely crazy ideas.”
Hoefler and his team received several awards and honors at SC22, including the Reproducibility Advancement Award for HammingMesh, a network topology for large-scale deep learning. Deep learning is a type of machine learning and artificial intelligence (AI) that imitates the way humans gain certain types of knowledge. Deep learning is an important element of data science, which includes statistics and predictive modeling. This new topology provides high bandwidth at low cost as well as high job scheduling flexibility.
Hoefler and his team also received an ACM SIGHPC Certificate of Appreciation for “Productivity, Portability, Performance: Data-Centric Python,” in which the team presents a workflow that retains Python’s high productivity while achieving portable performance across different architectures. Python is a high-level, general-purpose programming language. Its design philosophy emphasizes code readability with the use of significant indentation.
About the IEEE Computer Society Sidney Fernbach Award
Dr. Sidney Fernbach
Established in 1992 in memory of Sidney Fernbach, one of the pioneers in the development and application of high-performance computers for the solution of large computational problems. Dr. Fernbach is internationally recognized as one of the most influential scientists affecting the design of high performance computers during the period that the modern electronic computer grew from embryo to adolescence. From the first supercomputer systems, such as the Univac I and the IBM 704, to supercomputers a million times faster, Dr. Fernbach played a pivotal role.
Established in 1988, the annual SC conference has grown steadily in size and impact over time, drawing more than 13,000 attendees over the last several years. SC is held in a different city in the United States each year. SC has built a diverse community of participants including researchers, scientists, application developers, computing center staff and management, computing industry staff, agency program managers, journalists, and congressional staffers. This diversity is one of the conference’s strengths, making it a yearly “must attend” forum for stakeholders throughout the technical computing community. Indiana University has had a booth at the conference since 1997.
A snapshot from the Indiana University booth at SC22 in Dallas, Texas.