LOS ALAMITOS, Calif., Oct. 6, 2022 – The IEEE Computer Society has named ETH Zurich professor, Torsten Hoefler as the recipient of the 2022 Sidney Fernbach Memorial Award for his pioneering contributions to large-scale parallel processing systems and supercomputers.
Established in 1992 in memory of high-performance computing pioneer Sidney Fernbach, the Sidney Fernbach Memorial Award recognizes outstanding contributions in the application of high-performance computers using innovative approaches.
The ideas and software professor Hoefler and his group developed are actively used by tens of thousands of scientists today to power large-scale scientific simulations and artificial intelligence systems. Hoefler has been cited “for application-aware design of HPC algorithms, systems and architectures, and transformative impact on scientific computing and industry.”
Heading up ETH Zurich’s Scalable Parallel Computing Lab, Hoefler and his team have been instrumental in developing techniques to improve the efficiency of high-performance computing and large cloud data center systems. Many of the developed ideas form core components for constructing, running, and programming supercomputers such as the Alps system set to come online at the Swiss National Supercomputing Centre (CSCS) in spring 2023.
“One key architectural feature that distinguishes standard computers and commodity clouds with modern supercomputers that consist of millions of tightly coupled processors is networking, and Prof. Hoefler has made numerous innovative, groundbreaking contributions to enable high performance and programmability in such machines”, said Satoshi Matsuoka, director or Riken R-CCS and Professor at Tokyo Institute of Technology, the recipient of the 2014 Sidney Fernbach award, “In fact his contributions have been comprehensive, from work on innovative and scalable network topologies, various network routing algorithms, performance modeling, to making key contributions to the MPI standard, on which practically every scalable parallel codes are programmed upon. As scalability becomes significant in nontraditional HPC areas such as deep learning, Prof. Hoefler’s work will continue to be increasingly impactful for HPC and ultimately to the society.”
Hoefler’s work includes a focus on improving the performance of highly scalable parallel systems and developing numerous applications in the areas of weather and climate simulation, and machine learning. His contributions range from hardware and systems design for HPC to middleware, applications, as well as algorithms.
David Keyes, Director of the Extreme Computing Research Center at KAUST and 2007 Sidney Fernbach Award recipient, commented: “Torsten’s contributions to the analysis of parallel algorithms in terms of I/O complexity and data movement influence the approach of many scientists and, hopefully, will lead to a new understanding of algorithm design techniques for the next generation. He works not only on theory, but also on practical implementations, both aspects being important for continued scaling progress. Torsten also sets the standard, not only in terms of the software itself, but also in how performance is measured and reported, and even in how rigorous reviews should be conducted. He is an interdisciplinary integrator. His influence across many forms of computation has been truly transformative.”
Specifically, his contributions to the Message Passing Interface had a significant impact on computing.
Jack Dongarra, 2003 Sidney Fernbach awardee and winner of the 2021 Turing Award said: “Torsten played a significant role in shaping the Message Passing Interface (MPI-3) specification, the de-facto standard for programming distributed memory clusters and supercomputers. As an author of both the Collective Communication as well as Process Topologies chapters, he introduced key concepts such as non-blocking collective communication operations. His ideas have been adopted far beyond MPI as non-blocking all-reduce operations and are used in tens of thousands of deep-learning workloads. His reference implementation found its way into most open-source and commercial MPI implementations. His code is used on essentially all clusters and supercomputers by tens of thousands of researchers and practitioners.”
Hoefler has won numerous awards for his work, including being named as an IEEE Fellow and a member of Academia Europaea. He also received the Latsis Prize from ETH Zurich, the ACM Gordon Bell Prize, and two European Research Council (ERC) Grants.
His life work accelerates scientific research and applications of neural networks, making them more efficient and less costly. The impact of which will unlock new computational capabilities that enable breakthrough technologies in modern science.
“I am genuinely grateful to my mentors and the scientific community for their support and encouragement. I hope to continue Sidney Fernbach’s innovative path by combining computer architecture and applications,” said Hoefler.
With an eye on the future Hoefler anticipates tremendous growth, but also some of the challenges society faces in developing efficient large-scale deep learning training and small-scale inference systems. He believes that learning methods will have a fundamental impact on traditional simulation sciences such as weather and climate predictions, topics in which he is deeply invested.
The Sidney Fernbach award consists of a certificate and a $2,000 honorarium. The award will be presented to Hoefler at the SC22 Conference awards opening session in Dallas, Texas on Tuesday morning, November 15, 2022. He will also present an invited talk during the plenary session scheduled from 8:30 – 10:00am on Wednesday, November 16.
About the IEEE Computer Society
The IEEE Computer Society is the world’s home for computer science, engineering, and technology. A global leader in providing access to computer science research, analysis, and information, the IEEE Computer Society offers a comprehensive array of unmatched products, services, and opportunities for individuals at all stages of their professional careers. Known as the premier organization that empowers the people who drive technology, the IEEE Computer Society offers international conferences, peer-reviewed publications, a unique digital library, and training programs.
Source: IEEE Computer Society