Quantum machines currently have very little computing power. It’s still proving to be quite difficult to increase it. Researchers have recently proposed a new architecture for a universal quantum computer that goes beyond these constraints and may soon serve as the foundation for the next wave of quantum computers.
In a quantum computer, quantum bits (qubits) act as both a memory and a processing unit simultaneously. Because quantum information cannot be replicated, it cannot be stored in a memory like that of a conventional computer. This restriction necessitates that all qubits in a quantum computer be able to communicate with one another. This continues to be a significant obstacle in the development of potent quantum computers. In order to address this issue, theoretical physicist Wolfgang Lechner, along with Philipp Hauke and Peter Zoller, suggested a novel architecture for a quantum computer in 2015. This architecture is now known as the LHZ architecture after the authors.
Originally, according to Wolfgang Lechner of the Department of Theoretical Physics at the University of Innsbruck in Austria, “this architecture was created for optimization issues.” “In the process, we minimized the architecture to handle these optimization difficulties as effectively as feasible,” the author writes. The physical qubits in this architecture encode the relative coordination between the bits rather than representing specific bits individually. According to Wolfgang Lechner, this means that not every qubit is required to interact with one another. He and his team have now shown that this idea of “parity” is good for a universal quantum computer.
Complex operations are simplified
On a single qubit, parity computers can conduct operations between two or more qubits. According to Michael Fellner from Wolfgang Lechner’s team, “Such operations are already very well implemented on a modest scale by existing quantum computers.” However, the complexity of implementing these gate operations rises as the number of qubits does. Innsbruck researchers have now demonstrated that parity computers can, for instance, perform quantum Fourier transformations—a fundamental component of many quantum algorithms—with noticeably fewer computation steps and thus more quickly. Their findings were published in two articles in Physical Review Letters and Physical Review A. Because of our architecture’s strong parallelism, Fellner continues, “For example, the well-known Shor method for factoring integers may be done very efficiently.”
Two-stage error correction
The novel idea also provides hardware-efficient mistake correction. Quantum computers must constantly rectify faults because quantum systems are extremely sensitive to perturbations. The need for many more qubits arises from the necessity of devoting substantial resources to the protection of quantum information. According to Anette Messinger and Kilian Ender, two other members of the Innsbruck research team, “Our model runs with a two-stage error correction, one sort of error (bit flip error or phase error) being prohibited by the hardware utilized.” On several platforms, there are already preliminary experimental methods for this. According to Messinger and Ender, “the other type of problem can be recognized and remedied using the software.” This would make it possible to create the next generation of universal quantum computers with reasonable effort. Wolfgang Lechner and Magdalena Hauser’s spin-off company, ParityQC, is already working in Innsbruck with partners from business and science on potential applications of the new paradigm.
The Austrian Science Fund (FWF) and the Austrian Research Promotion Agency (FFG) provided financial support for the study conducted at the University of Innsbruck.
