New approach dramatically improves quantum chips
For many years, superconducting qubits were in the lead. Now a research team claims to have far surpassed previous successes - with quantum chips based on rubidium atoms.
So far, quantum computers are still at the stage of Albert Einstein shortly after he started school. The potential is there, but there is no guarantee that the child will one day become a revolutionary whose calculations will usher in a new era for physics. This is because the futuristic machines still make far too many mistakes. A research team from Harvard University, the Massachusetts Institute of Technology (MIT), the National Institute of Standards and Technology, the start-up QuEra and the University of Maryland has now presented a quantum processor with 280 qubits that can detect and correct errors in the scientific journal "Nature". To this end, they tested various error correction methods and executed complex, error-corrected quantum algorithms on 48 logical qubits. This work represents important progress towards a universally applicable quantum computer.
Qubits are the computing units of a quantum computer and the quantum mechanical equivalent of classical bits. A logical qubit, which performs the actual arithmetic operations, consists of several physical qubits. These are necessary to correct the errors that occur during quantum mechanical calculations. Most recently, the two US tech companies Google and IBM have set the pace in the development of ever larger quantum processors. For example, IBM presented a new chip called "Condor" with 1121 physical qubits on 4 December 2023. However, the sheer number of qubits installed is not the decisive factor. Many teams are therefore focussing on exploring new approaches and increasing the number of interconnected, logical qubits.
While Google and IBM use tiny superconducting circuits as qubits, in which electrical charges oscillate sometimes in one direction and sometimes in the other, the team led by Dolev Bluvstein, first author of the new "Nature" study, is focussing on excited states in rubidium atoms. To do this, the outer electrons are brought to very high energy levels (so-called Rydberg states) far away from the atomic nucleus using laser light - but they are not split off as in ions, so the atoms remain uncharged. In order to use them for quantum calculations, the researchers have to hold the excited atoms with additional lasers as if with tweezers. In this way, they can also be moved as required in a two-dimensional arrangement. The respective positioning of the qubits makes it possible to programme the machine. The advantage: unlike superconducting qubits, for example, the quantum chips do not have to be cooled down to a few millikelvin using helium; instead, the computers can be operated at room temperature. In addition, all atoms are identical and are not affected by any manufacturing inaccuracies. Various research groups in Germany are also working on such a quantum computer architecture.
Experts expect that the machines will one day be able to perform tasks that traditional computers fail at. For example, they could help in materials research, in the development of new medicines or in solving complex problems in the banking and insurance sectors. However, qubits are highly sensitive to external influences and often change their state unintentionally during the calculation. As a result, they repeatedly produce incorrect results. Recognising and correcting this without destroying the fragile quantum state is the task of error correction methods. Without such techniques, quantum computers cannot realise the potential attributed to them.
Quantum information cannot simply be copied
The particular challenge in contrast to the error correction mechanisms of classical computers, however, is the "no cloning theorem". It states that quantum information cannot simply be copied. This means that there is no "back-up". Nor does quantum mechanics allow the state of a single qubit to be read out without disrupting the entire calculation. To circumvent this problem, the stored information must be transferred from one qubit to an entangled system of several other qubits. The US computer scientist Peter Shor developed the idea for this back in the 1990s. Such a functional unit consisting of several physical qubits is known as a logical qubit.
Until now, it was assumed that an error-corrected logical qubit required more than 1000 physical qubits. A machine that can perform useful calculations would then have to have millions of physical qubits. However, Bluvstein's group has now succeeded in further improving previous error correction methods. How efficient the error correction is in the end is reflected in a figure called "code distance": larger code distances mean greater resistance to quantum errors, but also require more physical qubits. No other team in the world has yet demonstrated a code distance of seven. This enables the detection and correction of any three errors within a logical qubit, which can occur in each of the physical qubits. In addition, the team claims to have shown for the first time that increasing the code distance actually reduces the error tolerance in logical operations. This is not self-evident, as the probability of errors increases as the number of qubits increases.
The external reviewers of the research article describe the team's work as "impressive", as "significant progress" and as "proof that this technology has caught up and is now one of the leading architectures for a quantum computer". Immanuel Bloch, Director of the Max Planck Institute of Quantum Optics in Garching near Munich and himself one of the leading researchers in the field, described the results to "Spektrum.de" as "very nice work" that has a lot of potential.
The next step, however, must be to detect and correct the errors during the calculation, not afterwards. Only then would the system really come closer to being a universally applicable machine. "This is an exciting time for our field of research, as the quantum error correction and fault tolerance approach is beginning to bear fruit," said Mikhail Lukin, co-director of the Harvard Quantum Initiative and co-founder of the resulting start-up QuEra Computing. The progress has been achieved with a system that resembles an as yet unreleased second-generation quantum processor from QuEra, but is still installed on the premises of Harvard University. "This new step will significantly accelerate the development of powerful quantum computers and promote the next phase of innovation." Perhaps the quantum computer will receive a high school recommendation in a few years' time.
Spectrum of science
We are partners of Spektrum der Wissenschaft and want to make well-founded information more accessible to you. Follow Spektrum der Wissenschaft if you like the articles.
[[small:]]
Cover image: Shutterstock / Bartlomiej K. Wroblewski
Experts from science and research report on the latest findings in their fields – competent, authentic and comprehensible.