idw – Informationsdienst Wissenschaft
Nachrichten, Termine, Experten
Nachrichten, Termine, Experten
idw - Informationsdienst
Wissenschaft
In what order does a package delivery service route its deliveries to minimize fuel consumption and time? This type of problem – known as a combinatorial optimization problem – lies at the heart of many challenges in science, technology, and business. A new German-Taiwanese research project involving TU Darmstadt and National Cheng Kung University (NCKU) aims to solve such problems in the future using specialized computer chips, making the process significantly faster and more energy-efficient than has been possible to date. TU Darmstadt is receiving nearly 1.2 million euros in funding for this project from the German Federal Ministry of Education and Research.
Classic optimization problems – such as the famous “Traveling Salesman Problem,” the optimal partitioning of networks (Maxcut), or channel assignment in 5G networks – belong to the class of so-called NP-complete problems. This means that as the problem size increases, the computational effort required for classical computers rises exponentially, so that even high-performance computers eventually reach their limits. To make such problems more tangible, they can be represented as graphs and formulated using the so-called Ising model – a physical model that can be used to determine the energetically most favorable, i.e., “best,” state of a system. “Ising machines” offer a promising way to quickly find this minimum: these are special analog computers in which the phase of electronic oscillators maps the individual states of a graph, while weighted couplings between the oscillators represent its edges.
The challenge: With today’s chip technology, the necessary dense yet extensive interconnections are virtually impossible to implement. This is precisely where the German-Taiwanese research project “MesMerIsing” from TU Darmstadt and National Cheng Kung University comes in – “Memristor-based Machine for Rapidly Solving Optimization Problems in the Ising Form.” The researchers are relying on memristors – electronic components that, unlike classical transistors, can “remember” past states analogously and permanently. Applied directly on the chip, they could serve as fast, space- and energy-efficient coupling elements that also function at room temperature – without the need for complex cooling systems, such as those required by quantum computers.
The results of the research project – in which TU professors Klaus Hofmann and Christian Hochberger from the Department of Electrical Engineering and Information Technology, as well as Lambert Alff from the Department of Materials- and Geosciences (all part of the “Matter and Materials” research field), are participating – are relevant in the long term for numerous application areas: financial applications, drug research, supply chains and logistics, as well as AI and machine learning methods or classical image recognition. The chosen approach – specialized hardware instead of traditional processors – promises speeds several orders of magnitude higher while consuming significantly less energy.
The project strengthens the joint International Joint Research Lab (IJRL) “Memristor Technology” between TU Darmstadt and NCKU. The collaboration with Taiwan, one of the world’s leading semiconductor hubs, also bolsters the development of new microelectronics as a key technology in line with the German federal government’s high-tech agenda.
“MesMerIsing” is funded by the Federal Ministry of Research, Technology and Space (BMFTR) as part of research and innovation cooperation with Taiwan in the field of artificial intelligence.
TU Darmstadt will receive approximately 1.195 million euros over a three-year period for this project; the Taiwanese partners will be funded in parallel and separately by the National Science and Technology Council (NSTC).
About TU Darmstadt
TU Darmstadt is one of Germany’s leading technical universities and a synonym for excellent, relevant research. We are crucially shaping global transformations – from the energy transition via Industry 4.0 to artificial intelligence – with outstanding insights and forward-looking study opportunities. TU Darmstadt pools its cutting-edge research in three fields: Energy and Environment (E+E), Information and Intelligence (I+I), Matter and Materials (M+M). Our problem-based interdisciplinarity as well as our productive interaction with society, business and politics generate progress towards sustainable development worldwide.
Since we were founded in 1877, we have been one of Germany’s most international universities; as a European technical university, we are developing a trans-European campus in the network, “Unite!”. With our partners in the alliance of Rhine-Main universities – Goethe University Frankfurt and Johannes Gutenberg University Mainz – we further the development of the metropolitan region Frankfurt-Rhine-Main as a globally attractive science location.
MI No. 27e / bjb
Merkmale dieser Pressemitteilung:
Journalisten, Wissenschaftler
Elektrotechnik, Informationstechnik, Mathematik, Werkstoffwissenschaften
überregional
Forschungsprojekte, Wissenschaftspolitik
Englisch
Sie können Suchbegriffe mit und, oder und / oder nicht verknüpfen, z. B. Philo nicht logie.
Verknüpfungen können Sie mit Klammern voneinander trennen, z. B. (Philo nicht logie) oder (Psycho und logie).
Zusammenhängende Worte werden als Wortgruppe gesucht, wenn Sie sie in Anführungsstriche setzen, z. B. „Bundesrepublik Deutschland“.
Die Erweiterte Suche können Sie auch nutzen, ohne Suchbegriffe einzugeben. Sie orientiert sich dann an den Kriterien, die Sie ausgewählt haben (z. B. nach dem Land oder dem Sachgebiet).
Haben Sie in einer Kategorie kein Kriterium ausgewählt, wird die gesamte Kategorie durchsucht (z.B. alle Sachgebiete oder alle Länder).
