New Heads to Help Develop the Computers of the Future at Jülich
Jülich, 7 September 2021 - This summer, three top scientists moved from sunny California to Forschungszentrum Jülich. The researchers, who come from Google, Hewlett-Packard Labs and the University of California, will help to further expand the internationally outstanding research into new computer technologies at Jülich. The primary mission of the newcomers will be to put concepts for quantum computers and neuromorphic computers into practice.
The three newly appointed directors at the Peter Grünberg Institute are all focused on new forms of computing that are poised to fundamentally change the IT world. Quantum computers and neuromorphic computers already are the subject of intensive research at Jülich. The new approaches offer a lot of potential for certain tasks, in which conventional computers are increasingly being brought to their limits: for example in the field of artificial intelligence (AI) or in the case of complex simulations in materials research and medicine or the intelligent control of traffic flows.
On the road to quantum computers
At Google, Rami Barends was involved in developing the hardware for the first quantum computer that is demonstrably superior to a classical supercomputer. Now, he is continuing his work in the laboratories of Forschungszentrum Jülich. The 40-year-old physicist already has plenty of practical experience with quantum systems, most recently as a researcher and project manager in Google's Quantum Artificial Intelligence Lab at the renowned University of California in Santa Barbara. His switch to Jülich also brings the Dutchman closer to his home town of Delft where he completed his doctoral degree after graduating.
Quantum computing expert Prof. Rami Barends
Copyright: Forschungszentrum Jülich / Ralf-Uwe Limbach
At Jülich, he now heads the construction of a low-temperature laboratory where among others the quantum computers of the European Quantum Flagship project will be operated. Over the next few years, the physicist, together with his Jülich colleagues and external partners, wants to work on turning applications for quantum computers into a reality as soon as possible and on advancing the development of an experimental quantum processor made in Germany.
Putting neuromorphic computers into practice
The work of John Paul Strachan and Emre Neftci, on the other hand, is focused on neuromorphic systems that work similar to the human brain. Research on this topic at Forschungszentrum Jülich covers a worldwide unique spectrum - ranging from the fundamentals of materials science to computer architectures and algorithms to neuroscience. Neuromorphic computers could make AI applications significantly faster, more energy-efficient and robust. Moreover, they might help neuroscientists to gain a better understanding of the biological learning process than is possible using conventional computers.
Prof. John Paul Strachan researches neuromorphic systems
Copyright: Forschungszentrum Jülich / Ralf-Uwe Limbach
At Hewlett Packard Laboratories in Silicon Valley, John Paul Strachan most recently led a team researching neuromorphic hardware. The physicist and engineer, who was born in Costa Rica, currently holds more than 50 patents and previously studied and completed his doctoral degree at two leading American universities, MIT and Stanford. At Jülich, the 42-year-old wants to explore new brain-inspired computing concepts, build tangible hardware in the lab, and to help speed-up many scientific computing problems such as understanding how millions to billions of neurons interact with each other.
Prof. Emre Neftci explores neuromorphic software
Copyright: Forschungszentrum Jülich / Ralf-Uwe Limbach
Emre Neftci's field of research is closely related to this. The 39-year-old Swiss-American came to Jülich after several years of research at ETH Zurich (Switzerland) and the University of California in San Diego and Irvine (USA) to program applications for neuromorphic hardware with his team. Their focus is on redesigning machine learning algorithms from the perspective of neuroscience and realizing them in dedicated neuromorphic hardware. Because one thing is certain: the advantages of the new technology can only be fully exploited with the right algorithms and software.
