Quantum computing without mistakes is now possible

Artist impression of gate operations on logical quantum bits, that are protected from faults by means of quantum error correction. Credit: Johannes Knünz

Errors in how information is processed and stored are rare in modern computers because they are made so well. Error-correcting mechanisms that use redundant data are still used, though, for critical applications where even a single mistake can cause a lot of trouble.

Quantum computers are much more sensitive to disturbances, so they will probably always need ways to fix mistakes. If they don’t, mistakes will spread through the system and information will be lost. Quantum mechanics’ basic rules say that you can’t copy quantum information, so you can get redundancy by putting logical quantum information into an entangled state with several physical systems, like multiple atoms.

The team led by Thomas Monz of the Department of Experimental Physics at the University of Innsbruck and Markus Müller of RWTH Aachen University and Forschungszentrum Jülich in Germany has now realised for the first time a set of computational operations on two logical quantum bits that can be used to do any operation. Lukas Postler, an experimental physicist from Innsbruck, says, “For a quantum computer that works in the real world, we need a universal set of gates that can be used to programme any algorithm.”

Fundamental quantum operation realised

This universal gate set was put to use on an ion trap quantum computer with 16 trapped atoms by a group of researchers. Two logical quantum bits, each made up of seven atoms, were used to store the quantum information.

Now, for the first time, it has been possible to implement two computational gates on these fault-tolerant quantum bits, which are needed for a universal set of gates: a computational operation on two quantum bits (a CNOT gate) and a logical T gate, which is especially hard to implement on fault-tolerant quantum bits.

Markus Müller, a theoretical physicist, says that T gates are very basic operations. “They are especially interesting because quantum algorithms without T gates are easy to simulate on regular computers, which cancels out any speedup that might be possible. For algorithms with T gates, this is no longer possible.” The physicists showed how the T-gate works by putting a logical quantum bit in a special state and teleporting that state to another quantum bit using an entangled gate operation.

Fundamental building blocks for fault-tolerant quantum computing demonstrated. Credit: Uni Innsbruck/Harald Ritsch

The quantum information that is stored in encoded logical quantum bits is safe from mistakes. But this is useless without computations, and computations are also prone to mistakes.

Researchers have set up operations on the logical qubits so that errors caused by the physical operations underneath can also be found and fixed. So, they have made the first implementation of a universal set of gates on encoded logical quantum bits that can work even if something goes wrong.

“The implementation that can handle mistakes needs more steps than the one that can’t. This will cause more mistakes on the level of a single atom, but the experimental operations on the logical qubits are still better than logical operations that can’t handle mistakes “Thomas Monz is happy to say this. “It takes more work and is more complicated, but the end result is better.” The researchers also used numerical simulations on traditional computers to check and confirm the results of their experiments.

Now, the physicists have shown that a quantum computer has all the parts it needs to be able to work even when something goes wrong. Now, the goal is to use these methods on quantum computers that are bigger and thus more useful. The methods shown in Innsbruck on a quantum computer with an ion trap can also be used on other types of quantum computer architectures.

The study was written up in Nature.

Further information: Philipp Schindler et al, Demonstration of fault-tolerant universal quantum gate operations, Nature (2022). DOI: 10.1038/s41586-022-04721-1. www.nature.com/articles/s41586-022-04721-1

Journal information: Nature

Source: University of Innsbruck