
Useful quantum computers are one step closer. Microsoft and Atom Computing announced that they have set a new record for the most entangled logical qubits, which are crucial for developing quantum computers that can correct their own errors.
“We are co-designing and building the world’s most powerful quantum machine,” says at Microsoft.
Over the past decade, quantum computers have rapidly increased in size, there are more and more experts who can program them and they are becoming a viable tool for fundamental research. But despite this progress, quantum computers have not yet fulfilled the promise of solving problems that stump even the world’s best supercomputers.
Advertisement
Leaders in both academia and the burgeoning quantum computing industry have identified the devices’ propensity for making errors as the culprit – and the focus of much research has now shifted to building quantum computers that can catch and correct errors during computation. One promising approach is to take a quantum computer’s basic ingredients, called qubits, and connect them into groups of so-called logical qubits, as it is easier to correct errors when these groups are used for computations.
In 2023, a team of researchers at Harvard University and the quantum start-up QuEra set the record for the most logical qubits ever created in a quantum computer. These were made from extremely cold electrically neutral atoms of rubidium. Now, Atom Computing and Microsoft have used a quantum computer constructed from ultracold ytterbium atoms to create 24 logical qubits and link them through quantum entanglement.
The companies optimised the way lasers and electromagnetic fields control the atoms and encode information into their quantum states to turn them into qubits. Microsoft provided a “qubit virtualization” system that pinpoints the best way to group physical qubits into logical qubits, and Atom Computing provided the hardware.
“To my knowledge, this is the largest entangled state of logical qubits,” says Svore. She says this opens the door to using logical qubits to run algorithms, including those for error correction. In fact, to create this entangled state, the team already successfully employed some error-correction techniques, she says. In preliminary tests, using these logical qubits for simple calculations had an error rate about four times lower than when the team used conventional qubits.
“This is an impressive set of results that appear to advance the state-of-the-art in quantum computing with logical qubits encoded in neutral atoms,” says at the University of Wisconsin-Madison.
There is currently no consensus on the best way to design a quantum computer, but these results suggest that ultracold neutral atoms are promising. Companies like IBM and Google use qubits made from tiny superconducting circuits, while others rely on charged atoms or particles of light – and they are all racing to find the best way of implementing error correction.
at the California Institute of Technology says that such rapid progress in this type of quantum computing within just a few years “indicates that error correction with thousands, if not tens of thousands, of physical qubits is a near-term prospect”. Achieving this would put a whole new kind of problems within the reach of quantum computers and usher in a new era of searching for quantum advantage.
Svore and her colleagues are already working towards creating 50 logical qubits, and eventually want to create 100 of them, which they estimate would be sufficient to achieve real, practical breakthroughs in using quantum computers for materials science or chemistry.
arXiv