Digital “demon”: a method that increases the accuracy of quantum computing by 20 times

Researchers at UNSW Sydney have developed a new technique to reboot quantum computers with high precision. This process, known as preparing a quantum bit to a “0” state, is crucial for accurate quantum computing. The method is based on the principle of “Maxwell’s demon”, a hypothetical creature that can separate hot and cold molecules by observing their velocities. This innovative solution is simple yet effective in ensuring the reliability of quantum computing.

“Here we used a much more advanced ‘demon’ – a fast digital voltmeter – to observe the temperature of an electron randomly picked up from a hot electron beam. Professor Andrea Morello from UNSW, who led the team, said: “By doing this, we make it much colder than the pool it came from. “We’ve brought in a ‘0’ computational state, which is consistent with high confidence that the pool is in a computational state.”

“Quantum computers are only useful when they can reach a final result with a very low probability of error. And you can have nearly perfect quantum operations, but if the computation starts with the wrong code, the result is also wrong. Our digital ‘Maxwell devil’ tells us how precisely we can set the start of the computation to be the most useful. It gives more than 20 times.”

Professor Morello’s team pioneered the use of electron spins in silicon to encode and process quantum information, demonstrating record high precision, i.e. very low probability of error, when performing quantum operations. The final hurdle to efficient quantum computing with electrons was the accuracy of preparing the electron in a state known as the starting point of the computation.

Experiment author Dr. “The usual way to prepare the quantum state of an electron is to go to extremely low temperatures near absolute zero and hope that all electrons relax into the low-energy ‘0’ state,” explains Mark Johnson. “Unfortunately, even if we used the most powerful refrigerators, we had a 20 percent chance of accidentally priming an electron to the ‘1’ state. It was unacceptable, we should have done better.”

UNSW electrical engineering graduate Dr Johnson decided to use a very fast digital measuring instrument to ‘observe’ the state of an electron and use the device’s real-time decision processor to decide whether to store that electron and use it for further calculations. The result of this process was to reduce the probability of error from 20 percent to 1 percent.

A new twist on an old idea

“As we started to write down our results and consider how best to explain them, we realized that what we were doing was putting a modern spin on the old idea of ​​’Maxwell’s demon,'” says Professor Morello.

The concept of “Maxwell’s demon” dates back to 1867, when James Clerk Maxwell imagined a being that could know the speed of each gas molecule. He would buy a gas-filled box with a split center and a quick-opening door. Knowing the speed of each molecule, the demon can open the door so that slow (cold) molecules gather on one side and fast (hot) molecules on the other.

“The demon was a thought experiment discussing the possibility of violating the second law of thermodynamics, but of course such a demon never existed,” says Professor Morello.

“Now, we kind of created it using fast digital electronics. We just told it to observe an electron and make it as cold as possible. Here, ‘cold’ directly means that the quantum computer we want to build and run is in the ‘0’ state.”

The implications of this result are crucial to the viability of quantum computers. Such a machine can be built with the ability to make some mistakes, but only if they are rare enough. A typical margin of error is around 1 percent. This applies to all errors, including preparation, execution and reading of the final result.

This electronic version of Maxwell’s demon enabled the UNSW team to reduce training errors by twenty times, from 20 percent to 1 percent.

Dr. “Using a state-of-the-art electronic tool without the additional complexity of quantum hardware, we were able to prepare our electronic quantum bits with sufficient precision to provide reliable downstream computation,” says Johnson. “This is an important result for the future of quantum computing. And surprisingly, it’s also the embodiment of a 150-year-old idea!”