Dr. Valtteri Lahtinen Prof. Mikko Möttönen
CSO & Co-founder Co-founder and Chief Scientist at IQM
As stated in our previous article, different approaches are being taken to achieve a fault-tolerant quantum computer.
This article will look at one of the most popular approaches: superconducting qubits. We will discuss the advantages of this approach that make it the go-to method for companies like Google, IBM, and IQM.
We are also looking at problems standing in the way and methods of getting around them. By the end, we will better understand what superconducting qubits are and why they might be the way to go.
The idea of superconducting qubits is at least several decades old. It got more traction at the turn of the century when it was demonstrated that they are, in fact, physically possible.
It was shown that the 0 and 1 states could be measured, as well as the superposition between them. Even though the qubits back then were pretty lousy, maintaining their superposition for only a brief period, it was a significant achievement that created a lot of interest around the approach.
Superconducting qubits, at their very core, are electric circuits. There are different types of superconducting qubits that all have the same fundamental characteristics - the usage of superconducting materials being one of them.
Layout of IBM's five superconducting quantum bit device from 2015. (Credit: IBM Research)
Another common feature is the utilization of so-called Josephson junctions. These are points in the superconducting material that do not allow the classical flow of electricity through them. However, due to the phenomenon of quantum-mechanical tunneling, electricity does flow through the junction in tiny amounts.
As these junctions resist the flow of electric current depending on the total amount of the current, the junction's behavior is nonlinear.
In an electric circuit, the inductance and capacitance usually do not depend on the oscillation amplitude, whereas with the Josephson junction, this is the case. This means that the value of the inductance has a nonlinear dependence on the amplitude of the oscillation, creating a nonlinear electrical circuit.
This enables differentiating between different electrical states characterizing the 0, 1, and 2, as the properties of the circuit change during the transition of one state to another.
Importantly, we can measure these changes between the states.
One of the advantages of working with superconducting qubits is that we can design the electrical circuits to carry out the kind of dynamics we want.
This is possible due to the fact that we can mass-produce circuit elements allowing us to test different designs to see what works best.
This process is one of the key drivers of the steady advancements made in the circuits' quality and design. In fact, we might currently be relatively close to the point where the design itself is sufficient, allowing us to expand to a larger number of qubits.
This advantage, though, doesn’t come without a flip side.
When mass-producing, no one circuit is similar at the atomic scale, which causes differences in the behavior of circuits and decoherence in the qubits.
Furthermore, the usage of bulk materials carries some faults as these materials can contain impurities that then cause more decoherence in the qubits.
Currently, this is just the price to pay for being able to test with different designs and compositions.
There are many ways to face this problem, though, the most simple of them being better materials and enhancing the manufacturing processes. Likewise, new designs can also do their part in minimizing decoherence, and this is where modeling and simulations are key. For example, accurate electromagnetic simulations over a wide range of length scales are pivotal.
Decoherence will, however, always be prevalent. Therefore the realistic goal is not to eliminate the decoherence but to instead minimize it so that it doesn’t restrict the qubits performance excessively. At Quanscient, we are building simulation technology that can be one of the key enabling factors to achieve this.
When it comes to choosing the right path or the correct physical implementation for qubits, no one knows the right answer.
To seek guidance, we may look at history to see superconducting qubits offering steady development for more than 20 years. Compared to other implementations, superconducting qubits seem to have a more reputable background that could give hope for future advancements.
But as stated, before any technology is proven to provide useful quantum advantage, no one knows for sure.