Getting ready for quantum computing
Experts estimate that it will be at least another ten years before a quantum computer solves everyday problems. But large companies are already preparing to use the new computers. Forschungszentrum Jülich and the Helmholtz Research Field Information are supporting them in this endeavor. (Source: Forschungszentrum Jülich – Press Releases)
They are sitting at their fireplace, over which a stew is simmering, and are already working on a cookbook for their electric stove – which, however, first has to be designed and built. The situation is currently similar in quantum computing. The hardware is still in the midst of development, but researchers are already working on computational rules that a quantum computer should execute step by step in order to solve certain tasks, i.e. the quantum algorithms. Prof. Frank Wilhelm-Mauch, director of the Peter Grünberg Institute for Quantum Computer Analytics (PGI-12), believes that the simultaneous development of hardware and software makes sense and is important: “If you only start working on quantum algorithms when mature quantum computers already exist, you’ve lost a lot of valuable time.” Users who run these quantum algorithms on real quantum computers gain expertise and contribute to the co-design feedback loop. In this process, familiar from high-performance computing, users, software and hardware developers work together to improve and design future computing environments, clarifies Prof. Kristel Michielsen of JSC and head of the quantum computing user infrastructure JUNIQ.
Industry also sees a need for action: “Waiting until quantum technologies have reached mainstream status – that is, are a completely established technology – does not produce competitive advantages. And significant new markets will already be occupied by then,” warns the German information and telecommunications industry association Bitkom in its guide “Quantum Technologies in Companies.” Large corporations have therefore long since begun to build up competencies. One important task in order to be ready for takeoff when the first devices are ready: to find out today for which problems a quantum computer would actually have a practical benefit and what the computational approach – i.e., the algorithms – for solving such problems might look like.
To this end, science and industry are working closely together. Group leader Dr. Tobias Stollenwerk from PGI-12 believes that this is a win-win situation: “When we cooperate with industry, we learn to understand the problems that are important for companies. This helps us basic researchers to develop quantum algorithms not just out of scientific curiosity, but specifically to advance society with them.” He, his team and researchers at the Jülich Supercomputing Centre (JSC) are working on solutions for the automotive industry in the Q(AI)2 project together with BMW, Mercedes-Benz, Volkswagen and Bosch. The focus is on tasks in which the companies rely on artificial intelligence (AI) methods – so far using conventional computers.
One example: an AI is to recognize workers and their body parts in pictures. This is often the prerequisite for robots and humans to work together directly and safely during the production of a car part. The computer can then control the robot arm, for example, in such a way that a collision with a worker is avoided. Dr. Dmytro Nabok of JSC is exploring quantum algorithms to improve this AI application so that body part identification is always correct and within milliseconds.
To test the algorithms, Nabok is using, among other things, the quantum annealer made by D-Wave Systems at Forschungszentrum Jülich, called JUPSI. Unlike other quantum computers, a quantum annealer is not universally programmable, so it is only suitable for special tasks. The machine is part of the “Jülich User Infrastructure for Quantum Computers” (JUNIQ), through which research teams from industry and academia can access experimental systems, prototypes and commercial quantum computers even without Jülich participation. In addition, JUNIQ provides support in a quantum computer simulation laboratory, algorithm development groups and cooperative research.
Volkswagen AG is dealing with another exemplary problem in the automotive industry. The initial situation: New cars are to be painted in two separate layers. When a car arrives at the paint shop, however, the robots there do not apply the two coats of paint directly one after the other. This is because a large number of cars arrive on the conveyor belt and the robots would then have to constantly change the paint coats. This involves time-consuming replacement and cleaning of the painting tools. For this reason, a belt conveys each car out of the system after the first color coat has been applied and then back in again later for the application of the second color coat. The optimizer’s task is now to determine the timing of a color change for a certain number of cars so that it has to be done as rarely as possible. Tobias Stollenwerk’s team is investigating the extent to which quantum computers can solve such problems faster than conventional computers. In particular, they are shedding light on how the real-world limitations of real quantum hardware affect the performance of these algorithms.
“The computational cost of this problem increases tremendously fast with the number of paints and the number of cars. In addition, other influencing factors have to be taken into account in reality,” Stollenwerk explains. “Such tasks sometimes cannot be solved exactly at all in realistic time spans, not even with supercomputers.” Quantum algorithms should help greatly speed up optimization problems like these using quantum computers.
Unlike Nabok, Stollenwerk’s team tests the quantum algorithms they develop on conventional computers that mimic the behavior of quantum computers. “This allows us to specifically co-simulate different types of errors that quantum computers of different maturity levels make. This helps us better understand the impact of these errors on the outcome,” Stollenwerk says.
JUNIQ project manager Dr. Nils Küchler reports a pleasant side effect: “Quantum computing has also given conventional computers an advantage by developing quantum-inspired software.” Stollenwerk has observed another phenomenon while working with industry: “Companies shed new light on problems when they consider how quantum computers might compute them. In the process, they sometimes discover far better classical algorithms than those they were previously aware of.”
FZJ/B. Schunk, 14.04.2023
The original press release can be found at:
Fit werden für den Quantencomputer (only in german)
Localization in the Helmholtz Research Field Information:
Helmholtz Research Field Information, Program 2: Natural, Artificial and Cognitive Information Processing, Topic 2: Quantum Computing
Contact:
Prof. Dr. Frank Wilhelm-Mauch
Peter Grünberg Institut
Institute for Quantum Computing Analytics (PGI-12)
Forschungszentrum Jülich
Phone: +49 2461 61-6106
E-Mail: f.wilhelm-mauch@fz-juelich.de
Prof. Dr. Kirstel Michielsen
Quantum Information Processing
Jülich Supercomputing Centre (JSC)
Forschungszentrum Jülich
Phone: +49 2461/61-2524
E-Mail: k.michielsen@fz-juelich.de
Dr. Tobias Stollenwerk
Peter Grünberg Institute (PGI)
Institute for Quantum Computing Analytics (PGI-12)
Forschungszentrum Jülich
Phone: +49 2461 61-2076
E-Mail: to.stollenwerk@fz-juelich.de
Dr. Nils Küchler
Project manager „Jülich UNified Infrastructure for Quantum computing“ JUNIQ
Institute for Advanced Simulation (IAS)
Jülich Supercomputing Centre (JSC)
Forschungszentrum Jülich
Phone: +49 2461 61-6877
E-Mail: n.kuechler@fz-juelich.de
Dr. Dmytro Nabok
Institute for Advanced Simulation (IAS)
Jülich Supercomputing Centre (JSC)
Forschungszentrum Jülich
Phone: +49 2461 61-6579
E-Mail: d.nabok@fz-juelich.de
Contact for this press release:
Dr. Barbara Schunk
Press Officer
Forschungszentrum Jülich
Phone: +49 2461 61-8031
E-Mail: b.schunk@fz-juelich.de
