Error correction can lead to usable quantum computers – New Scientist

It has been shown for the first time that it is possible to reduce the total number of errors of a quantum computer. That means we can build bigger, usable quantum computers.

Quantum error correction is an important step for the development of usable quantum computers. Now Google has shown that its approach to error correction is scalable. Researchers of the company report this in the scientific journal Nature. This step provides confidence that practical quantum computers will appear on the market in the coming years.

The building blocks of a quantum computer are qubits, the quantum variant of the transistors in a classical computer chip. But qubits are prone to errors. Those need to be detected and corrected if we want to build quantum computers large enough to tackle real-world problems.

surface code

A popular approach to error correction is called surface code correction. Many qubits work together as one so-called logical qubit. This provides redundancy: if one qubit makes a mistake, the combined whole continues to work. This is also how error correction works in classical computers.

In quantum computers, however, there is an additional complication. Any attempt to measure qubits immediately destroys the data.

This means that adding more physical qubits to your logical qubit can also be detrimental. “When engineers try to organize larger ensembles of physical qubits into logical qubits, with the aim of making fewer mistakes, the opposite has actually happened so far,” says quantum computer expert Hartmut Neven from Google.

Google showed this when it first presented a working error correction scheme in 2021, which resulted in a net increase in errors. Researchers at the Joint Quantum Institute in Maryland later succeeded in creating logic qubits that did not further worsen error rates, albeit in a rather technical way, rather than a practical one.

Commercially usable quantum computer

Now Google has shown that logic qubits can be expanded and that this scaling reduces the overall error rate. If that trend can continue, they can perform calculations that would be impossible even on the most powerful classical computers. Neven says the team now has “palpable confidence” that Google can make a commercially useful quantum computer.

Raster bits
The team reached the milestone using the third generation of Google’s Sycamore quantum processor, with 53 qubits. surface code logical qubits are typically a grid of qubits linked to other qubits. One qubit is always reserved to measure the value of other qubits. The company’s experiment switched from 3-by-3 grids, with 17 physical qubits, to 5-by-5 ​​grids with 49 qubits. That means that almost the entire processor works as one logical qubit. This increase reduced the margin of error from 3.028 percent to 2.914 percent.

Scaling up

Google’s team admits the improvement is small, but says the scaling process can theoretically continue indefinitely. This paves the way for a quantum computer that makes few mistakes and can therefore perform useful tasks reliably. But the next step, moving to a logical qubit of 6 by 6, which requires 71 physical qubits, is impossible with the current generation of quantum processors. That requires better hardware.

Quantum information researcher Fernando Gonzalez-Zalba of the University of Cambridge says a greater improvement in the margin of error would have been nice, but the research is moving in the right direction. “The individual components in the processor still need to improve a bit to increase the logic error margin as the technology scales,” he says, “but what we see in the team’s releases is that they improve significantly with each release. I don’t think scalable quantum error correction will take years, I think we’re almost there.’