In late 2019 it was hard to not stumble upon Quantum Computing. But, let’s quickly recap what has been happening a few months ago. Google announced it has achieved “Quantum Supremacy”, claiming that its 53 Qubit chip Sycamore took only 200 seconds to solve a simulation task that would take 10.000 years on a today’s super-computer.
This was immediately followed by criticism by IBM explaining that current super-computers could indeed solve the task that was presented if used efficiently. However, they came up with an estimate of 2.5 days - which is still about a factor 1000 slower than the quantum computer. In conclusion, quantum computing competes with the best super-computers - however, it has not yet opened an entire new era…. yet. We’ll stay tuned to see that. In Germany the federal government has decided to support Quantum Computing and is currently evaluating potential facilities to host a respective competence center. Also, in Austria there is a respective hotspot in Tyrol.
The question is now, what’s behind this new technology that can enable so much for us. On the lowest layer, we need to understand that quantum particles show some properties and effects that contradict our classical understanding of physics and are thus not really intuitive. The most important of these effects is superposition. It means that a particle can be in several states at the same time. Think of a piano where the played tone is its state. You can also have multiple tones being played at the same time resulting in a respective superposition in the sound wave. The second important quantum property is entanglement. This means that two spatially distinct particles have a common state.
Fine so far? But how can we use that to build the fastests computers ever? In a next step, we introduce Qubits. Instead of “classical” bits being either 0 or 1, these quantum bits can also have several states at once. When we describe this in a mathematical way, this is similar to using complex numbers with an additional imaginary part instead of real numbers. When really physically building such a machine, instead of observing if the electric potential is above or below a threshold we need a representation that can have another dimension as well. An example for this is observing the spin-vector of an electron that can point upwards or downwards but also any direction in between.
To do computations on Qubits or quantum registers as combinations of several Qubits, let’s introduce quantum gates. However, in contrast to classical electronic gates that are building blocks within an electrical circuit, a quantum gate is rather a manipulation on a same quantum register. This can be physically implemented by, for example, applying a magnetic field. A classical circuit then becomes a fixed set of manipulations being applied after each other.
So much for physics… But how does this help to solve complex computations more efficiently than classical computer architectures? Stay tuned for our next story on how to actually programme a quantum computer.