Quantum annealing with more than one hundred qubits (1304.4595v2)

Abstract: Quantum technology is maturing to the point where quantum devices, such as quantum communication systems, quantum random number generators and quantum simulators, may be built with capabilities exceeding classical computers. A quantum annealer, in particular, solves hard optimisation problems by evolving a known initial configuration at non-zero temperature towards the ground state of a Hamiltonian encoding a given problem. Here, we present results from experiments on a 108 qubit D-Wave One device based on superconducting flux qubits. The strong correlations between the device and a simulated quantum annealer, in contrast with weak correlations between the device and classical annealing or classical spin dynamics, demonstrate that the device performs quantum annealing. We find additional evidence for quantum annealing in the form of small-gap avoided level crossings characterizing the hard problems. To assess the computational power of the device we compare it to optimised classical algorithms.

• The paper demonstrates that the D-Wave One device operates via genuine quantum annealing, evidenced by its distinctly bimodal success probability distribution.
• It uses simulated quantum annealing to reveal close energy and success correlations, sharply contrasting with the results from classical annealing.
• The study highlights small-gapped avoided level crossings as key quantum features, suggesting further research into scalability and hybrid optimization methods.

Analysis of Quantum Annealing with a 108-Qubit D-Wave One Device

The paper presented in this paper explores the domain of quantum annealing using a 108-qubit D-Wave One device. It provides experimental data supporting the operation of the D-Wave device as a quantum annealer, in strong contrast to classical annealing dynamics. The focus is on the performance and behavior of quantum annealing as compared to classical optimization approaches, particularly in the context of solving NP-hard problems such as the Ising spin glass model.

Key Findings and Analysis

The research successfully demonstrates that the D-Wave One device essentially performs quantum annealing by analyzing the correlation between the performance of the device and simulated quantum annealing (SQA). The correlations are significantly more aligned than those between the device and simulated classical annealing (SA). This indicates that quantum mechanics, particularly quantum tunneling, plays a key role in the optimization process on this device. The paper highlights several noteworthy observations:

• Bimodal Success Probability: The D-Wave device exhibits a distinct bimodal distribution of success probabilities in identifying ground states, a pattern also seen in SQA. In contrast, SA results displayed a unimodal distribution, hinting at the fundamentally different mechanisms at play between quantum and classical approaches.
• Energy-Success Correlation: The research reveals correlations not only in success probabilities but also in energy distributions of final states. The D-Wave device and SQA show closely resembling distributions, implying the presence of quantum characteristics in the device's operation.
• Avoided Level Crossings: Instances deemed difficult for the D-Wave device are associated with small-gapped avoided level crossings, which are characteristic of quantum processes. This highlights quantum coherence in traversing energy landscapes where such phenomena are critical.

Implications and Future Directions

The results imply that quantum annealing, as executed in the D-Wave One, can process optimization problems with different efficiencies compared to classical methods. This observation reinforces the potential of quantum annealers in solving certain complex computational problems more efficiently than classical devices. However, the paper stops short of claiming exponential quantum speedup. Instead, it posits that improvements in coherence times and qubit count could eventually lead to a quantum advantage over classical methods.

Looking toward the future, these findings suggest several research pathways:

• Device Scaling: With more qubits and enhanced qubit coherence, larger and more difficult problem instances could be tackled, potentially showcasing more explicit evidence of quantum speedup.
• Algorithmic Enhancements: Combining quantum annealing with classical post-processing could refine results and exploit both paradigms' strengths.
• Benchmarking and Comparative Analyses: More comprehensive comparisons across a broader range of problem classes could clarify the specific domains where quantum annealers excel.

Theoretical and practical studies should be directed toward understanding the scalability of quantum annealing and exploring hybrid approaches that incorporate quantum and classical computational techniques. Such synergy could unlock new levels of performance enhancement in combinatorial optimization and beyond.

In conclusion, the conducted experiments underscore the quantum mechanical nature of the D-Wave processes, providing steps towards leveraging quantum technologies in specialized computational tasks. Future research focusing on device enhancements, problem-specific developments, and operational benchmarking could significantly advance the field of quantum computing, particularly quantum annealing.