Microsoft researchers say they have created “topological qubits”, long sought-after components for a radically different kind of quantum computer. This isn’t the first time the firm has made this claim – it attempted to produce these error-proof quantum bits in a similar experiment in 2023, but the results weren’t fully conclusive, raising doubts among colleagues in the field about whether it has fully worked this time.
Topological qubits could solve one of quantum computing’s biggest problems: all existing quantum computers make too many errors. Many researchers are working to develop quantum machines that can correct their own errors as they compute. But Microsoft has championed a different approach. Its researchers have focused on creating qubits – the basic building blocks of a quantum computer – that would have special, error-preventing properties from the start.
Now, they say they have made several such qubits and created a method for using them in computations. The firm claims to have built an eight-qubit chip, named Majorana 1, but it has yet to release any performance data.
Advertisement
At the core of all Microsoft’s topological devices are superconducting nanowires layered on top of the semiconductor indium arsenide. When each of these wires superconducts, both of their ends see the emergence of a special quasiparticle – a coordinated behaviour of many other particles such as electrons – called a Majorana zero mode (MZM).
Encoding information into pairs of these quasiparticles and combining two wires would turn them into a qubit. Quantum properties of MZMs, such as them being their own antiparticle, would make those qubits exceptionally error-proof.
Furthering their previous work with MZMs, the researchers tested a method of reading out the quantum state of each quasiparticle pair by connecting the wire that contains them to three quantum dots, or tiny semiconductor crystals. at Microsoft says that past work gave his team over 90 per cent certainty that it had detected MZMs, and being able to read out their state significantly strengthens that case.
Majorana 1, with its reported eight topological qubits made of four MZMs each, relies on this, as the presence of MZMs is critical for making it work. In their latest paper, which focuses on the quantum dot method, the researchers write that “this measurement does not unequivocally distinguish” between MZMs and other possibilities, but they estimate the probability of incorrect interpretation to be 1 per cent.
The new work comes in the wake of several scientific controversies. In 2021, a paper involving the company’s researchers was after another team of physicists imperfections in the device could have mimicked signatures of MZMs. Two years later, another paper reporting evidence of MZMs came under fire due to similar concerns about data interpretation, with researchers outside the firm saying that Microsoft didn’t share enough data for others to replicate the findings.
at Delft University of Technology in the Netherlands says the method employing quantum dots to measure the quantum states of MZMs is a necessary building block for a topological quantum computer, but the new work doesn’t provide definitive evidence that the quasiparticles were present in the device.
at the University of St Andrews in the UK says that for this measurement to be meaningful, it would also have to be repeated on many more devices than Microsoft has presented so far.
Akhmerov and Legg both say that the quantum dot procedure relies on a previous method developed by the company that isn’t impervious to mistaking imperfections in the superconducting wires for the elusive quasiparticles.
“Scepticism is healthy in science,” says Nayak, but his hope is that each new measurement and experiment will serve to diminish it, making it harder to come up with plausible alternative explanations for the findings.
at the California Institute of Technology says that the quantum dot method the team created is meaningful even if what has been measured are “Majorana impostors”.
“The pursuit of a topological qubit is a very, very worthwhile endeavour. But in the topological qubit land, we’re still most certainly at the zero-qubit level. No one has ever made one, but at least what Microsoft has demonstrated is a crucial ingredient that you would need to work your way up,” he says.
Nayak’s view of the topological qubit landscape is more optimistic and he is confident enough in his team having created several MZM-based qubits to keep pushing ahead. The team is already working on creating and testing devices with more than eight qubits.
Nature
