UNSW Sydney engineers have riffed on the famous Schrödinger’s cat analogy to demonstrate a more efficient way to eliminate errors in quantum computing.
“Imagine you’re trying to find your cat hiding in one of eight identical cardboard boxes, in a dark and noisy room,” says UNSW Scientia Professor Andrea Morello.
“You are not allowed to enter the room – opening the door may kill the cat . What is the optimal strategy to find out where it’s hiding? Our team of quantum researchers have found an answer to this problem, and it might be an important milestone on the road to building a quantum computer.”
The cat metaphor has been used for decades to illustrate the quirky phenomena that occur when attempting to apply quantum mechanics to macroscopic systems.
In a UNSW-led study the “cat” is the nucleus of an atom of antimony, implanted in a silicon quantum chip. Although atomically small, the antimony nucleus has eight quantum states that can be used to encode quantum information.
Having eight different states leaves extra room to detect and correct errors that may occur during the calculation.
Improving quantum error correction is a major hurdle in building large-scale quantum computers.
“The quantum states used to encode the information are called, indeed, “ Schrödinger cat ” states,” says Prof. Morello.
“The key remaining problem is finding out an error has occurred, without disrupting the precious information encoded in the atom – or ‘cat’.”
Just add ‘water’
To explain how this problem was solved, Prof. Morello says to imagine the cat in the dark room.
“In this scenario, you cannot enter the room and look inside, so instead you could place eight sprinklers in the room, each placed above one of the boxes. You then spray some water over each box, in sequence, and listen for an angry ‘meow’ when the cat expresses displeasure for being sprayed.
“But because the room is noisy, you might mistakenly think that a meow came from an empty box – or miss registering a true meow coming from the box containing the cat.”
He says the standard method to reduce the chance of such mistakes is to repeat the whole experiment several times, and infer the cat is in the box where the most meows came from.
However, spray too often and the cat might panic and jump to another box.
“Repeatedly sprinkling the boxes risks changing the very thing you are trying to observe,” Prof. Morello says.
The trick, says Prof. Morello, is to stop immediately once the first “meow” is observed – this is your initial guess – and switch to sprinkling only the boxes where the cat supposedly isn’t.
“Silence from those boxes increases the confidence that your guess was correct,” he says.
“The absence of a signal confirms the presence of another, without interacting directly with the system.
“Sometimes, silence can be loud.”
Back to the science
In the physical system used by the researchers, the ‘sprinkler’ is an electron that can be pushed onto the atom and then removed from it conditionally on the quantum state of the nuclear spin.
The addition and removal of the electron can unsettle the nucleus and make it jump to a different state.
When the new strategy is applied to the ‘atomic cat’, the electron only needs to come off the atom once. After that, only the empty states are probed.
This method more than halved the chance of error, and it cut the total measurement time to a third .
Lead author Arjen Vaartjes says using this adaptive measurement strategy, the team managed to boost the confidence of ‘finding the cat in the right box’ to 99.61%.
“This value is significant because it puts our system in the range of measurement fidelities necessary to perform successful quantum error correction.”
Prof. Morello says, “quantum error correction relies on repeated measurements without disrupting the fragile quantum information, equivalent to finding the cat in the right box without scaring it.”
By changing their measurement strategy, the team showed it is possible to extract more information while causing less disturbance – an essential step toward utility-scale quantum computing.
The team says the approach could improve ‘mid-circuit’ measurements used in quantum error correction, which is a major challenge in developing scalable quantum computers for applications such as drug discovery, simulation of chemical reactions, optimisation of financial portfolios and machine learning.
A trick any quantum lab can use
Prof. Morello says the broader impact of the work is that it can be applied to a wide range of other quantum computing systems.
“This adaptive measurement approach may help significantly reduce measurement errors in systems ranging from semiconductor qubits to atomic or photonic architectures,” he says.
“Because many architectures also employ similar hardware, the new protocol can readily be adapted to other platforms that suffer from errors during measurement.”
He says scalable quantum computing may ultimately depend on how well we learn to ‘find the cat in the right box’ without disturbing it.
“We can now extract information about the quantum system just gently enough to keep it intact.”
Vaartjes says “all it took was a fast FPGA, a cup of coffee, a dedicated team of clever researchers and a long Friday afternoon of coding”.
PRX Quantum
Maximizing the Nondemolition Nature of a Quantum Measurement Via an Adaptive Readout Protocol