Quantum Computer Achieves Record 10⁻² Fidelity with 50 Qubits

Philipp Aumann and colleagues at the Parity Quantum Computing GmbH (Austria and Germany) and the Institute for Theoretical Physics at the University of Innsbruck, have achieved a record performance in fidelity and qubit count for the quantum Fourier transform, a key algorithm for many quantum applications. Using the Parity Architecture on the IBM Heron r3 chip, they attained a process fidelity of approximately 10^-2 for a 50-qubit quantum Fourier transform. This represents a super-exponential speedup compared to previous methods and suggests further improvements are possible with expanded instruction sets.

Quantum Fourier transform fidelity exceeds previous limits with 50 qubit demonstration

Process fidelity for the quantum Fourier transform (QFT) reached 10^-2 for 50 qubits, a substantial leap beyond the previous best result of approximately 10^-2 achieved for 36 qubits. This improvement is significant because the QFT is a foundational component in numerous quantum algorithms, including Shor’s algorithm for factoring large numbers and quantum phase estimation, both of which have implications for cryptography and materials science. The previously attainable fidelity levels for QFTs of comparable size were insufficient to yield reliable results, as the accumulation of errors during the computation would overwhelm the signal. This threshold of 10^-2 signifies a level of accuracy previously unattainable for computations of this scale, as earlier attempts suffered from error rates that rendered results unreliable. The demonstrated super-exponential scaling, specifically (exp(N²)), means performance improves dramatically with each additional qubit, unlike previous swap-based approaches which typically scale polynomially. This exponential scaling is crucial for achieving quantum advantage, where quantum computers can solve problems intractable for classical computers.

Incorporating iSWAP gates into the instruction set further improves scaling, enabling new approaches to complex quantum calculations. iSWAP gates facilitate the exchange of quantum information between qubits without requiring physical movement, reducing the overhead associated with long-range interactions. Experiments utilising XX dynamical decoupling, a technique to suppress noise and decoherence, and layouting by the qiskit transpiler achieved record performance in both fidelity and qubit count. Dynamical decoupling involves applying a series of carefully timed pulses to the qubits to average out the effects of environmental noise. While these fidelity levels represent an advance, they do not yet guarantee error-free computation for truly complex problems; further improvements are needed to overcome limitations in current quantum hardware, such as qubit coherence times and gate errors. The iSWAP-based Parity Twine method requires only one third of the gate count compared to prior algorithms, contributing to this efficiency. Reducing the gate count is vital as each gate operation introduces a potential source of error.

Achieving high-fidelity quantum Fourier transforms via parity-based compilation

Parity Twine Networks emerged as key to this advancement, offering a systematic way to compile complex quantum algorithms like the QFT, a mathematical operation akin to converting a sound wave into its component frequencies, but performed on quantum data. The QFT is essential for transforming a time-domain signal into its frequency-domain representation, and its quantum implementation offers potential speedups over classical algorithms. These networks tackle the challenge of needing widespread qubit connections on processors that naturally favour nearest-neighbour interactions. Most current quantum processors have limited connectivity, meaning that qubits can only directly interact with their immediate neighbours. This necessitates the use of SWAP gates to move quantum information around the chip, increasing circuit complexity and introducing errors. Instead of physically rearranging qubit states using iSWAP gates, this approach operates on ‘parity information’, encoding the relative alignment of logical variables within the computation. Parity refers to whether the number of ‘1’s in a binary string is even or odd, and this information can be used to represent and manipulate quantum states without requiring direct qubit interactions. This result represents a record performance in both fidelity and qubit count for this benchmark on a quantum processor utilising a CZ-based instruction set, achieved by reducing gate count and circuit depth through operation on parity information. The CZ gate (controlled-Z gate) is a fundamental two-qubit gate commonly available on many quantum processors.

Fifty qubit parity architecture demonstrates super-exponential speedup but faces algorithmic

Achieving this level of fidelity and qubit count represents a vital step towards realising the potential of quantum computation for tackling complex problems. The abstract offers limited insight into how readily this architecture scales beyond the QFT, a specific mathematical operation used as a benchmark. While the QFT is a valuable test case, the true challenge lies in demonstrating the scalability of this approach to more complex and application-specific algorithms. Previous work highlighted the challenges of maintaining performance as algorithms become more intricate, but acknowledging these concerns does not diminish this achievement. The complexity of quantum algorithms often leads to increased circuit depth and gate count, exacerbating the effects of noise and decoherence.

Reaching both high fidelity and a substantial qubit count of fifty is a landmark in quantum hardware development, particularly utilising readily available CZ-based instructions. This architecture demonstrably outperforms previous swap-based methods with a super-exponential speedup, with improvements via iSWAP gates further strengthening its potential. Such progress, even if focused initially, establishes a crucial foundation for broader quantum computation applications. The ability to perform complex calculations with a relatively small number of qubits is a significant advantage, as building and controlling large numbers of qubits remains a major technological hurdle.

This advance establishes a new standard for quantum computation by demonstrating a streamlined method for managing qubit interactions on physical hardware. A process fidelity of approximately one hundredth for fifty qubits during a quantum Fourier transform represents a significant leap forward. The resulting super-exponential scaling outperforms previous techniques reliant on rearranging qubit states, offering a pathway to more efficient and complex calculations. The Parity Architecture’s ability to reduce the need for SWAP gates is particularly noteworthy, as these gates are often a bottleneck in quantum circuit execution. Further research will focus on extending this architecture to other quantum algorithms and exploring its potential for solving real-world problems.

The research successfully demonstrated the Parity Architecture on quantum hardware, achieving a process fidelity of approximately 10⁻² for a quantum Fourier transform using fifty qubits. This represents a record performance in both fidelity and qubit count for processors utilising CZ-based instructions and offers a super-exponential speedup compared to previous swap-based methods. The authors indicate that future work will explore applying this architecture to other algorithms and problem-solving. This advance establishes a foundation for more efficient quantum computation by streamlining qubit interactions and reducing reliance on SWAP gates.