The company believes devices with a million topological qubits are possible.
Many years ago, Microsoft committed itself to an exotic, high-risk approach to quantum computing. Now, as the company unveils its first prototype chip, it says its perseverance has paid off.
One of the biggest problems bedeviling today’s quantum computers is their susceptibility to errors. Even the slightest interference from the outside world can collapse the fragile quantum states they rely on to carry out computations.
Quantum error correction provides a potential workaround, but the most promising schemes require huge numbers of extra qubits. As a result, most experts predict machines will need roughly one million qubits before they can do anything truly useful, which is a long way from today’s record of a little over 1,000.
That’s why nearly two decades ago Microsoft decided to pursue topological qubits—a novel type of qubit that is inherently resistant to errors. The effort has faced several setbacks over the years, but now the company claims it’s finally cracked the problem with a processor featuring eight topological qubits and will one day be able to host up to a million.
“We took a step back and said ‘OK, let’s invent the transistor for the quantum age,’” Chetan Nayak, the Microsoft technical fellow who led the effort, said in a press release.
What sets topological quantum computing apart from other approaches is that rather than encoding quantum information in individual particles, it encodes data in the macroscale properties of a larger system that can consist of many particles. It builds on the mathematics of topology, which focuses on the properties of objects that stay the same even if they are bent or stretched.
The main advantage of this approach is that disturbances to individual components of the system don’t affect its overall topological state. Topological qubits, therefore, are much less susceptible to the kind of environmental noise that causes errors in other kinds of qubits.
Microsoft’s approach to topological quantum computing involves creating so-called quasiparticles known as Majorana zero modes. It has done this by combining a nanowire made of the semiconductor indium arsenide with a plate of aluminum that acts as a superconductor at very low temperatures.
Normally electrons in superconductors pair up. But Microsoft says its device can generate unpaired electrons that exist in a “delocalized” state. These present as a pair of Majorana zero modes, one at each end of the nanowire.
If you can create four at either end of a pair of nanowires, it should be possible to “braid” them into a topological state to encode quantum information. This braiding process involves making a series of measurements in a specific order.
However, Microsoft has yet to provide convincing proof the approach actually works. The announcement of the new chip coincided with the publication of a paper in Nature describing experiments conducted on the new device.
But these simply outlined a way to measure whether or not Majorana zero modes exist in the nanowires by detecting if there are an odd or even number of electrons. This is a crucial step because these two states will effectively act as the 0s or 1s in the company’s qubits. But the paper doesn’t provide solid evidence that Majorana zero modes are present. They simply validate the measurement approach.
Nayak told MIT Technology Review that his team has unpublished results providing more definitive proof, which they are currently writing up into a paper. But there’s likely to be a certain amount of skepticism, as there has been controversy around Microsoft’s previous publications on this topic.
A 2021 paper in Nature reporting the detection of Majorana zero modes was later retracted after other physicists suggested the signatures could have come from defects in the device used to create them. Another paper claiming evidence of the quasiparticles in 2023 was also criticized for not providing enough information for other researchers to reproduce the results.
Nonetheless, the company is confident it has now cracked the topological qubit and is firmly on the path towards building a large-scale quantum computer in years rather than decades. It has also been keen to tout the fact that DARPA seems to agree. Microsoft is one of two companies that have made it through to the final phase of the agency’s competition to find unusual quantum approaches that could achieve practical scale much faster than conventional wisdom suggests is possible.
It’s likely to be some time before there’s consensus on the significance of Microsoft’s latest result. And the journey from a scientific demonstration like this to a practical product is long and fraught with risk. But if the company’s approach pays off, it could dramatically speed the advent of the quantum age.
The post Microsoft Claims Quantum Computing Breakthrough With Its Exotic New Chip appeared first on SingularityHub.
* This article was originally published at Singularity Hub
0 Comments