Quantum computing: Is it really all it’s cracked up to be?
Quantum computers could change our world. But not in the way you may think. For those working on this complex technology, managing expectations is proving a challenge. DW spoke to one of Europe’s top quantum scientists.
Imagine this — the quantum computer has taken over the world. Oldschool, classical computing is only talked about in university lecture halls. During every millisecond of every day, deep in dark basements, great quantum machines munch through numbers to support our artificially intelligent peers. A groundbreaking drug is developed weekly. Security codes have been cracked open. The quantum financial trader is on the loose.
This is what the “quantum age” could look like. Or so some would have us believe.
Google’s announcement in October that its quantum computer, Sycamore, had achieved “quantum supremacy” — meaning it carried out a calculation that a classical computer could not in a realistic amount of time — allowed imaginations to run wild.
Such a technological milestone was celebrated, challenged and stewed over. Immediately, people started asking: how could this change our world? Some proffered suggestions — many of which pushed the boundaries of reality, verging on science fiction.
Why not? One of the many joys of science is to think without limits. Nonetheless, for scientists working day in, day out, on quantum computing, it might be time to put expectations back into perspective.
Unlocking the mysteries
Quantum computing uses the properties of quantum effects as its foundation. It’s entirely different to classical computing, but can still be used alongside it. Quantum computers, when they are larger and function predictability, will help us develop new drugs and financial systems.
They will be a key to unlocking advances in artificial intelligence and cyber security. They will not break into your bank account and steal all your money. Once the science reaches that level, quantum encryptions will undoubtedly emerge.
Instead of “bits” — the currency of classical computers — quantum computers use quantum bits, or “qubits.” A classical bit can have the value 0 or 1. But a qubit can be both and everything in between. This is a quantum principle known as superposition.
When qubits are in a superposition they can become entangled, meaning they communicate with each other. Using these principles a quantum computer can tackle several tasks at once, whereas a traditional computer has to solve each problem one at a time. For example, a 10-qubit quantum computer can process 1024 possible inputs at once.
This means these machines could be much faster than the supercomputers we have today. We can’t assume, however, that they will lay waste to classical computers, because there could be cases where a quantum computer cannot compute what a supercomputer can.
Are you still with me? For non-physicists, quantum computers can be difficult to imagine — so to wrap my head around it, I went deep into the German countryside.
To the forest
The race to be the leader in quantum computing has already begun. Whoever masters this technology will be at an advantage. In the United States, IBM and Google lead the way. D-Wave in Canada is another powerful player. China is pouring vast sums of money into development and the European Commission is making quantum technologies a priority in the digital single market.
One of the key players on team Europe resides in a forest in western Germany. Not where you would imagine them to be, perhaps — but working on world-changing science needs to be removed from any noise. The Jülich Supercomputing Centre is buried in the Stetternicher Forest in Germany’s north west.
Getting there from the Deutsche Welle headquarters in Bonn involved a train to Cologne, another to Düren and a long wait for the rural train to Jülich Forschungszentrum. A journey that took around three hours. It would have taken an hour in a car but having just arrived from the UK, I wasn’t quite ready to take on the high speeds of the German autobahn while driving on the other side of the road.
Sitting in her office, Kristel Michielsen was welcoming and direct. Under her leadership, as a Quantum Information Processing professor, researchers at Jülich played a role in Google’s demonstration of quantum supremacy.
They used their supercomputer, called JUWELS, to simulate circuits with up to 43 qubits in order to benchmark the Google processor. For Professor Michielsen, quantum computing is a new, disruptive technology.
However, she’s keen to highlight that the gun is often jumped, saying, “A lot of promises are made and it’s not always stressed that it’s still in a kind of research phase.”
“Sometimes people think: Okay, it went fast with mobile phones, it went fast with this and that, so maybe in a few years I will have my own quantum computer in a mobile phone. I think this is simply not realistic. If we even have working quantum computers in the first stage, they will be in a computer center like this one,” she told DW.
A working quantum computer
Right now, you can’t turn on a quantum computer and have it work perfectly. Machines with fewer qubits will encounter fewer errors but there are still problems to be solved.
“If we switch on a conventional computer we know if we do a calculation today, and we do it tomorrow, we will get the same outcome,” says Michielsen.
“For the quantum computer, if we get an answer today, tomorrow it can be different, so we have to be aware of this. These errors are not just statistical errors, meaning if we repeated the computation many times we would improve our answer.”
The point is, Michielsen says, with these devices there are still systematic errors that the experts are currently unable to resolve. These machines are also extremely sensitive and any unexpected stimuli can hinder results.
In fact, there are many aspects of quantum computing that still remain a mystery.
For quantum scientists, it’s imperative that such mysteries are unravelled in order to make progress. “One cannot build a lot of applications on mysteries,” says Michielsen.
“Of course, many things in the past have been developed without us understanding the physics behind it.” She gives the example of the steam engine — which was invented without creators understanding every detail of the way it functioned. Similarly with quantum computing, Michielsen says, we don’t have to unlock all the secrets of this complex science to advance, but the more we know the better.
So what can quantum computers do today?
It depends on what kind of quantum computer you are talking about. The most common gate-based machine, the ones used by IBM and Google, don’t have a lot of applications — yet. The calculation that Sycamore carried out last month using 53 superconducting qubits isn’t useful in daily life. At the moment these machines can do very simple algorithms.
“We can add numbers,” says Professor Michielsen. “One can factorize small numbers — so that’s always this prime factorization. One can also simulate small molecules, but this isn’t reliably accurate.”
The other main type of quantum computer is a quantum annealer. As with D-Wave’s machine, annealing allows for more prototype applications. These machines have 2,000 qubits and work differently to the way gate-based computers operate, but still use quantum effects.
Using D-Wave’s technology, Volkswagen carried out research on traffic flow optimization in Beijing. Researchers at The Roswell Park Cancer Institute in Buffalo, New York, used quantum annealing to optimise the amount of intensity-modulated radiotherapy that could be given to two prostate cancer patients.In 2018, Qx Branch, a data analytics and quantum computing company, used quantum annealing for machine learning to predict the outcome of the 2016 US Presidential race.
What about in Europe?
The campus in Jülich is a collection of modern buildings with an industrial feel. The central canteen overlooks a picturesque lake. Overall, it’s understated.
The supercomputing centre is building up a laboratory for Europe’s first quantum computer that can compete with IBM and Google. Ultimately, Europe doesn’t want to be dependent on privately owned tech giants for quantum technologies.
OpenSuperQ is an EU initiative that aims to builda quantum computer with up to 100 superconducting qubits. It’s hoped it will be completed by September 2021 — after which it will be accessible to researchers within the bloc.
Jülich already has elements of a quantum computer, such as the chandelier and a chip, which I wasn’t able to see. With a nervous laugh, Professor Michielsen admitted she hadn’t seen them either.
Across Europe and the world, computer labs have small quantum computers with between one to 15 qubits. Normally these are prone to errors and don’t have the same technology as IBM, Google or D-Wave. Quantum trade secrets are closely held.
You can login and use a quantum computer via the cloud. The IBM Q Experience is a lesson in the complexities of working in this field. Using a dizzying array of coloured squares users can build complex circuits and put them on the machine. Thousands of people have already run experiments on IBM Q systems and simulations.
Quantum computer scientists are continually encouraging people to make use of this technology because the more that is put on the machines, and the more the machine gives us, the more we can learn about how they work.
The real quantum age
Quantum computing has come a long way. Google’s “quantum supremacy” achievement seemed impossible to many scientists even a year ago. Like a child’s first foray onto its two feet, every step is important and nothing should be taken for granted.
It may be that quantum computers will have to work closely with classical computers for some time. Nonetheless, Michielsen says that will still be beneficial.
“In the development of new drugs, if quantum computing couldn’t do the job completely alone, but helped in the development, then that will be a huge step.”
Quantum computing is part of the future. How different that future will be is anyone’s guess. Its impact could be more subtle than we imagine now — but, hey, even if it’s just traffic jams that are banished to the history books, we’ll have something to be thankful for.