Google’s quantum calculator was the pioneer – and maybe the first quantum machines will soon be practically usable. But which applications benefit from it? In this article, cityofmayfield.org tell their viewpoints.
The term quantum computing has haunted the IT scene for years. Scientists and practitioners hope that quantum machines will perform certain computing tasks many times faster than classical computers. So far, research on quantum computing has been more theoretical.
That changed suddenly in the fall of this year. On October 23, Google announced that it had reached “Quantum Supremacy”. This term refers to the moment when a quantum computer is superior to conventional computers for a specific task for the first time.
The Sycamore processor that Google’s researchers developed can solve a task that even the fastest supercomputer would take 10,000 years to accomplish in 200 seconds. This breakthrough is made possible by a complete departure from the way digital computers work.
Quantum computers work on the basis of quantum mechanical principles. While in classic computing a bit can only assume the state 0 or 1, the so-called QuBits can assume several states simultaneously in quantum computing. For a shorter period of time, states between 0 and 1 – that is the so-called superposition – are possible and then fall back to 0 and 1. Because of these completely new, quantum mechanical principles, such a computer is able to carry out significantly more arithmetic operations simultaneously.
How are quantum computers deployed?
However, the fact that Google shows the superiority of the quantum calculator based on a purely academic problem does not mean that a market launch is imminent. On the contrary: The application potential of quantum computers is difficult to predict today. Too many unresolved issues still need to be addressed. Pessimists expect a broader availability only in about 10 years, optimists speak of a few years.
Nevertheless, certain application and application scenarios are emerging. For example, quantum computers will certainly not be used on-premise on site. At most, they could expand the range of applications for PCs and other IT devices by using cloud services, writes the “Office for Technology Assessment” at the Bundestag in a brochure. The technical challenges for the functioning of quantum processors – such as massive cooling to an absolute zero of -273 degrees and operation in a vacuum chamber – are simply too demanding to be used locally at a company on site.
Quantum computers will also not replace traditional computers, but rather complement them. Commercially available computers have a significantly lower susceptibility to failure compared to quantum computers and are superior in many areas of application. According to the experts for technology assessment, data centers are likely to be created for quantum computers that can be accessed externally via a network.
The problem of the traveling salesman
As far as the tasks that can be practically solved are concerned, two types are particularly suitable for quantum computers: One type is the fast search in huge, unordered amounts of data. The second is the rapid “factorization” and areas in which the “combinatorial explosion” plays a role.
The standard example of the combinatorial explosion is the Traveling Salesman problem: A traveling trader wants to visit around 16 cities to sell his goods. In what order should he visit the cities to save as much time as possible? With 16 cities there are already more than a thousand billion possible route combinations – and with each city more the problem exponentially expands. According to the calculation by a computer scientist at the University of Trier, classic computers that try out all possible combinations take 20 years with the simplest algorithm and still 4.7 hours with an optimized process.
A quantum computer would – instead of trying out all possible combinations one after the other – calculate the travel time of all possible routes simultaneously. Even if one or more cities were included in the calculation, this increases the difficulty and the solution time for the quantum computer, but not exponentially as with the classic computer. The solution would be available in a split second.
Practical applications 1: From big data to material design
The example shows where the rabbit is going: Quantum computing has its greatest use wherever complex calculations are involved, in which a multitude of possibilities are considered. This results in the practical applications in which quantum computing unfolds its potential.
- Big data and databases: Companies that operate large databases and process huge amounts of data could benefit from quantum computing. A classic computer must look at each data point individually when searching large amounts of data. A quantum computer could speed up the search enormously and reduce the response time considerably.
- Traffic routing: Future mobility will require mastering highly complex processes that push current computer generations to their limits. If a traffic jam occurs, today’s systems only report traffic jams. The performance of a quantum computer can be used to calculate in real-time for each individual car when it is better to turn right or left in order to prevent the occurrence of a traffic jam in advance. VW has already shown in a research project that quantum computers can be used efficiently to control traffic.
- Material design and optimization: With the quantum computer, a company can, for example, develop new materials more efficiently in the manufacturing industry or optimize mixing ratios. VW, for example, is testing material optimization for batteries in electric vehicles, according to a Handelsblatt report. In the future, it should be possible to simulate molecular flows using quantum computers.
Practical applications 2: From genetic research to the financial industry
- Genetics and biology: With the computing power of quantum computers can tap into genetic research and biotechnology completely new possibilities. For example, penicillin is a relatively large molecule; no supercomputer is able to simulate a molecule of this size. A quantum computer, on the other hand, only requires 286 qubits to represent and simulate a molecule. Such simulations are of great value for the pharmaceutical industry because they can save material-intensive and expensive tests.
- Artificial intelligence: machine learning based on large amounts of data is much more efficient with quantum computing. Today’s deep learning – “deep neural networks” – is linked to hard combinatorial optimization problems. These can be solved much faster and better by quantum computers than by classic computers – which could make machines much smarter.
- Financial industry: Quantum computers can outperform traditional computers in the high-frequency trading of securities. Coupling quantum computers with AI can also provide more precise results for decision support in trading. Banks can also use quantum computers in the profiling of their customers to detect fraud or optimize investment portfolios.
There is a high probability that there will be many other possible applications that are not yet in view at all. So for now, wait and see. Prof. Kristel Michielsen, Head of Quantum Information Processing at Forschungszentrum Jülich, summed up the situation in a ZEIT interview: “I would compare Google’s quantum computer to a hot air balloon that is just rising: it is flying somewhere – we only know the exact direction not yet. But the real breakthrough is that the quantum balloon flies at all.”