The near-term future of quantum computers
You are a professor at the electrical and computer science department and QuTech. What are your main research interests?
I am interested in quantum error correction and its physical realizations, in particular the use of superconducting qubits. I also find the theory behind superconducting qubits, called circuit-QED, fascinating, as it is a novel, emerging, theory of quantized electric circuits. I have a long-standing interest in the power of quantum computers, that is, what you can and cannot do with them.
What are near-term applications of a quantum computer? Often the near-term future is called the “NISQ (noisy intermediate-scale quantum) era”. What does that mean and what kind of applications might be possible?
NISQ means that the current chips have a limited number of qubits (say 10-100 qubits), and logical operations on these qubits are inaccurate. This means that we can only do a limited number of operations on these qubits before the outcome of these operations becomes unreliable and therefore useless. One can program different types of quantum dynamics on these devices, for example aimed at simulating some small physical or chemical system, or solving a small quantum or classical optimization problem.
What role does the recent Google supremacy experiment play in this regard?
The Google supremacy experiment and follow-up experiments has shown that we are entering an era where the classical simulation of quantum experiments on NISQ devices becomes very costly and time-consuming. This by itself does not mean that the quantum device is solving an important or interesting problem, it just means that its different capability, which was theoretically understood previously, is starting to become an experimental reality.
...Quantum Inspire is useful in giving researchers who are not quantum physicists access to quantum chips, allowing them to familiarize themselves with how they function.
Most of the known promising quantum algorithms require very small errors when doing qubit operations. To this end error correction is essential (one of your main research topics). What is the current status of quantum computers in this regard? What are the main obstacles to achieve quantum error correction?
Quantum error correction experiments with 4-20 qubits are taking place and in these experiments one is looking at how well errors can be corrected. Quantum error correction works by adding redundancy. Nine physical qubits are used, say, to make one, better, so-called logical qubit. If error rates are low enough, then adding more redundancy helps, so one uses, say, 25 qubits to make an even better logical qubit. So the challenges are: first, are error rates low enough. And second, can we handle the redundancy, that is, can we build and control bigger chips with many more qubits (and still keep error rates low).
What is your personal expectation: what will be the first useful application of a quantum computer and when would it happen?
I expect that quantum computing will be useful for understanding a novel and common property of the dynamics of many-body correlated quantum systems that may have eluded us because of limitations of classical algorithms. Such application is not necessarily useful for society-at-large. Let’s hope we can achieve this goal in the NISQ area, say in the next 10 years, by simply building bigger chips without requiring the full realization of quantum error correction.