The Swiss Federal Institute of Technology (ETH Zurich) is a science and technology university with an outstanding research record. ETH Zurich is the study, research and work place of 18,000 people from 80 nations. About 350 professors in 15 departments teach mainly in the engineering sciences and architecture, system-oriented sciences, mathematics and natural sciences areas and carry out research that is highly values worldwide.
The Department of Information Technology and Electrical Engineering (D-ITET) includes a dozen research laboratories with activities in the broad area of electrical engineering, ranging from integrated circuits to computer networks, from wireless communications to signal processing, and from control theory to power electronics. The faculty of D-ITET consists of about 30 professors that supervise over 300 Ph.D. students. The department is firmly committed to both basic and applied research.
The Digital Circuits and Systems group led by Prof. Benini focuses on architecture, hardware and software design of energy-efficient computing systems. The group is involved in a number of EU funded research projects (e.g. HERCULES, PHIDIAS, PRO3D, VIRTICAL in the ICT Cooperation program, SCALOPES, SMECY in the ARTEMIS JTI, MODERN, END In the ENIAC JTI), including the ERC AdG-MULTITHERMAN project focusing on sustainable computing systems. The group is also very active in industrial cooperation and technology transfer with major companies in the computing field, such as Intel, NXP, ST Microelectronics and Samsung Electronics.
Contribution to OPRECOMP
The goal of OPRECOMP is to develop computation systems that perform over a very large operation range and achieve an energy efficiency that is simply unattainable with traditional digital design paradigms. Unless there is a major shift in the way systems are designed, performance of future systems will be at the mercy of technological developments forecast by Moore’s Law, which slowly and surely is reaching physical limits.
With this project we want to rethink the way we do computations and take inspiration from nature, where multiple very simple organisms with extremely limited capabilities come together to produce impressive feats of calculations. OPRECOMP will rely on small, fast and efficient systems to perform the bulk of the calculations with various degrees of accuracy. Unlike traditional approximate computing systems, that limit the accuracy of specific calculation units, in this project we plan to free data in the intermediate stages of the calculation from accuracy constraints not only during computation, but also while it is stored and transferred within the system without affecting the quality and validity of the final result.
Classical digital computing systems have always been optimized under the assumption that all calculations are accurate. Once this constraint is lifted, many interesting opportunities appear that allow to trade-off the performance with the accuracy. ETH Zürich will develop computation nodes that will exploit this trade-off and allow data to be processed, stored and transmitted at different accuracy levels while still maintaining accurate control flow of the system and enabling the system to precisely bound and control the accuracy of the final result.
Chair of digital Circuits and systems at ETHZ and a full professor at the University of Bologna
He has served as Chief Architect for the Platform2012/STHORM project in
STmicroelectronics, Grenoble in the period 2009-2013. He has held visiting and consulting researcher positions at EPFL, IMEC, Hewlett-Packard Laboratories, Stanford University.
Dr. Benini’s research interests are in energy-efficient system design and Multi-Core SoC design. He is also active in the area of energy-efficient smart sensors and sensor networks for biomedical and ambient intelligence applications. In these areas he has coordinated tens of funded projects, including an on-going ERC Advanced Grant on Multi-scale thermal management of Computing Systems.
He has published more than 700 papers in peer-reviewed international journals and conferences, four books and several book chapters (h-index=85 with more than 31,000 citations on Google Scholar). He is currently ranked 1st worldwide in Microsoft’s “Top computer Science Authors” in the field of “Hardware & Architecture” in the last 10 years.
He is a Fellow of the IEEE and a member of the Academia Europaea and has served for two terms as a member of the steering board of the ARTEMISIA European Association on Advanced Research & Technology for Embedded Intelligence and Systems.
Frank K. Gürkaynak
Frank K.Gürkaynak obtained his B.Sc. and M.Sc. degrees from the Istanbul Technical University, and his Ph.D. degree from the Integrated Systems Laboratory of ETH Zurich. Since 2008 he is with the Microelectronics Design Center of ETH Zurich, involved in research and teaching on the design and test of digital integrated circuits and is a member of the PULP project.
For the OPRECOMP project, he’s responsible of package 4.
Michael Schaffner received his BSc. and M.Sc. degrees from the Swiss Federal Institute of Technology Zurich, Switzerland, in 2009 and 2012. He is currently pursuing the Ph.D. degree with ETH Zurich. He has been a Research Assistant with the Integrated Systems Laboratory and with Disney Research Zurich, since 2012. His research interests include digital signal processing, video processing, and the design of very large scale integration circuits and systems. Michael Schaffner received the ETH Medal for his Diploma thesis in 2013.
Gianna Paulin received her B.Sc. and M.Sc. degree from the Swiss Federal Institute of Technology Zurich (ETHZ), Switzerland, where she started as a Ph.D. student at the beginning of 2019. Her main interests lay in deep learning and computer architecture.
In OPRECOMP, she is working on alternative number representations for Transprecision Computing.
Stefan Mach received his B.Sc. and M.Sc. degree from the Swiss Federal Institute of Technology Zurich (ETHZ), Switzerland, where he is currently pursuing a Ph.D. degree. Since 2017, he has been a research assistant with the Integrated Systems Laboratory at ETHZ.
In OPRECOMP, he is working on the design of hardware units for Transprecision Computing.
Fabian Schuiki received his B.Sc. and M.Sc. degree in electrical engineering from ETH Zürich, in 2014 and 2016, respectively. He is currently pursuing a Ph.D. degree with the Digital Circuits and Systems group of Luca Benini. His research interests include computer architecture, transprecision computing, as well as near- and in-memory processing.
For the OPRECOMP project, he’s doing System integration and RTL work on the kW platform, interfacing a PULP system via CAPI with a POWER8 machine.