Researchers can now start unleashing their algorithms on the quantum computer, on the platform Quantum Inspire. However, they’ll only have access to a digital replica of the computer – a digital twin. The physical computer isn’t ready for use yet, Kokkelmans says. “It doesn’t work well enough yet to make it available to the general public.” The computer was actually supposed to be ready as early as 2024, but due in part to the relocation of the labs to Qubit, flooding from heavy rainfall, and air conditioning problems, that didn’t work out.

Nevertheless, the decision was made to put the digital twin online this week. “We’re simulating the quantum computer on a very powerful ‘traditional’ computer. Its computing power scales exponentially for every qubit you add, to mimic a real quantum computer,” Kokkelmans says. Qubits are building blocks for quantum computers, similar to bits for traditional computers.

Atoms

In the TU/e computer, the qubits consist of atoms. “We have a lump of metal, which we evaporate, turning it into a gas. We make the atoms of that gas so cold that they can no longer move. Then we pluck them out of the cloud and we can put them wherever we want,” Kokkelmans explains. When those qubits are in the right place, you can perform operations on them.

Researchers, but also the general public, can now use the online interface on Quantum Inspire to have the computer execute a program, because it’s available to everyone who is interested. “The user enters a code, which prescribes how step-by-step operations take place. When the program is finished, the qubits are in a certain position. What you read out is the result of the calculation.”

Hybrid

The quantum computer in Qubit is a hybrid quantum computer, combining traditional and quantum computing technologies. The best of both worlds, is how Hamid Montazeri describes it. As an assistant professor at the Built Environment department, he will be one of the first to use the new computer. “We already have a solid foundation in classical computing. Quantum computers offer enormous potential for the most complex problems.” He will use an algorithm to hand the most complex parts of his calculations to the quantum part of the computer, the rest will go to the traditional part.

Quantum computers aren’t faster than traditional computers in all tasks, but in specific ones they excel. These include cracking encryptions, searching huge databases, and mimicking natural processes. Being a physicist, Kokkelmans personally finds the latter especially interesting. “Molecules, for example, have properties similar to a quantum computer. A molecular problem, like a quantum computer, grows exponentially with every atom you add. So with a quantum computer, you can properly calculate how molecules behave, because the computer behaves the same way.”

It’s very similar to how an ordinary computer works, but the big difference is in what the qubits do. They may, for example, be instructed to go into superposition, which means they simultaneously represent 0 and 1. You can also entangle qubits with each other, inextricably linking the state of one qubit with that of another.

Simultaneously

Superposition and entanglement are properties within quantum physics that are hard for many people to grasp, but that are real nonetheless. And they allow the computer to perform multiple calculations simultaneously. “With an ordinary computer, there’s a number and you perform an operation on it. Then the number changes and you perform another operation on that one,” Kokkelmans explains. In doing so, the computer builds on each previous step. “With a quantum computer, all the numbers are in there at the same time.”

As an example he mentions a telephone directory – for young readers: a large book containing the numbers and addresses of all the people in a certain area, arranged by last name. If you only have a phone number, it’s virtually impossible to find the name that goes with it by flipping through the pages. A traditional computer would take that approach, scanning the numbers until the right one was found. With a quantum computer, things are different. “It invokes the function for all numbers simultaneously. If it’s not the right number, then that number gets a 0; if it is the right number, then it gets a 1.” You only need a mere twenty qubits to call on a million numbers simultaneously.

For a quantum computer, it’s important that the qubits are stable, and therein lies a challenge, according to Kokkelmans. Sometimes a qubit gets ‘lost’, which doesn’t help the reliability of the computer. This is why, at the moment, it can’t be used for real applications, but that may change soon. “I hope that the people who are going to use the computer will use it to do things that we can’t do ourselves. We’re physicists. We create a facility and do understand a thing or two of algorithms, but I’m curious to see what others come up with.”

New possibilities

A group at the Built Environment, led by Hamid Montazeri, will make use of the computer. Kokkelmans: “With supercomputers, researchers are running into limitations: they can no longer use them to solve everything. This is a new direction that may give them that possibility in a few years’ time.”

In any case, says Kokkelmans, developments are moving at lightning speed. “For the past few years, we’ve been in an interesting phase where things are really happening.” Whether everyone will actually have a quantum computer on their desk at some point remains to be seen, but even about this the professor is increasingly optimistic. In future, he would like for chips to come equipped with different laser pulses for each qubit. “If we can pull that off, then I think it will be possible to make chips that can operate ten thousand atoms simultaneously,” he says. That development, he says, could become a reality in just a few years. Although... He adjusts the expected number of qubits down a bit: “For now, let’s say several hundreds.” Still, that’s a lot, compared to the nine that are available online now. But at least it's a start.

Fluid flow modeling

Hamid Montazeri’s work involves all kinds of flows: air flows inside ASML’s chip machines, for example, and the ones around wind turbines. The calculations for those kinds of flows are complex and contain a lot of details. To understand how gas, water, and air behave, researchers use computer models.

Montazeri shows a model, a visualization of what happens to the air around a rotating wind turbine. Colors indicate how fast the wind is moving and vortex patterns form around the blade of the mill. Such a model can take weeks to process, even with a supercomputer. And a model of an entire wind farm with multiple rotating wind turbines? Forget about it. Even the most powerful classical supercomputers aren't strong enough to create that kind of simulation.

Algorithms

That’s why Montazeri has been keeping a close eye on developments in quantum computing for years. Now his team will be the first to use TU/e’s computer. Creating full simulations won’t be possible yet, but the group will be able to test algorithms and investigate the most efficient way to use the quantum computer for this application.

The number of qubits to be used may be low now, but he has high hopes for the future. “Quantum hardware is getting better and better, so the software should gradually evolve along with it. The goal is to test our algorithms on the quantum computer, improving it at the same time. We have to work together to grow.” In any case, it’s clear to him that the algorithms for the quantum computer should make processing simulations many times faster. “How fast exactly depends on the complexity of the computation and on the maturity of both the hardware and the algorithms, but our early results are very promising.”