SWAPNET circuit boosts quantum state purity for near-term devices

Researchers at the Massachusetts Institute of Technology have resolved a two-decade-old problem in quantum purity amplification (QPA), offering a new strategy to combat the degradation of quantum states caused by noise. The team presents the optimal QPA protocol for all quantum systems and global noise, and introduces SWAPNET, a sparse and shallow circuit designed to enable QPA on near-term devices.

Under conditions of strong depolarization, the protocol achieves an exponential reduction in sample complexity compared to existing methods, potentially improving performance on Noisy Intermediate-Scale Quantum (NISQ) devices where resource limitations are significant. Simulations and experiments using superconducting quantum processors confirm the protocol’s robustness and practical utility.

NISQ Device Limitations Motivate Quantum Purity Amplification

This sparse and shallow circuit addresses a critical need for improving the reliability of quantum computations without relying on the resource-intensive methods of full fault-tolerant quantum computing. QPA offers a process-agnostic approach, requiring only access to output states rather than detailed knowledge of the computational process itself. By using multiple copies of noise-corrupted states, the protocol aims to generate a final state with higher purity, a key metric for successful quantum information processing.

Unlike quantum error correction, which demands substantial overhead in qubits and complex control sequences, QPA is particularly beneficial when algorithms can be rerun and results stored in quantum memory, but achieving high output fidelity remains a challenge. The protocol’s efficiency stems from its ability to reduce the number of samples needed to achieve a desired level of purity, a factor that becomes increasingly important as quantum systems scale.

Zhaoyi Li, Honghao Fu of the Computer Science and Artificial Intelligence Lab at and Concordia University, Takuya Isogawa of the Department of Nuclear Science and Engineering at, Caio Silva at, and Isaac Chuang at authored the study.

Optimal QPA Protocol Resolves Two-Decade Open Problem

The mathematical difficulty of optimizing quantum purity amplification (QPA) has been overcome by converting the problem into a symmetric form, building on recent advancements in combinatorics, and resulting in a three-step protocol of Schur sampling, correction, and trace-out. This resolution extends beyond depolarizing noise to establish optimal sample complexity in terms of infidelity for a wider range of inputs, using d-dimensional states, qudits, to provide a structured mathematical framework for the process.

The team’s work formally defines the optimization of QPA as a task, then presents its solution and operational interpretation. These results, detailed in Section IV of the paper, showcase applications in Hamiltonian simulations, adiabatic evolutions and state preparation on IBM superconducting quantum processors. Looking ahead, the team suggests that extending QPA to larger quantum systems with generic noise models and exploring its broader applicability will open new avenues for practical applications in quantum computing, particularly within the Noisy Intermediate-Scale Quantum (NISQ) era.

QPA’s Advantage Over Quantum Error Correction & Mitigation

Quantum purity amplification presents a distinct path for enhancing quantum computations, particularly for applications where obtaining multiple copies of a quantum state is feasible. This process-agnostic approach allows QPA to generate higher-purity states by consuming several copies of states degraded by noise, a benefit for tasks involving quantum simulation, sensing and machine learning.

The practicality of quantum error correction on current Noisy Intermediate-Scale Quantum (NISQ) devices is limited by the substantial overhead introduced by coding, with advantages only appearing at larger block sizes or with multiple layers of concatenation. Achieving universal quantum operations on logically protected qubits also demands complex methods, such as magic state distillation, and increases hardware requirements for platforms like superconducting circuits, trapped ions and neutral atoms.

QPA circumvents these challenges by directly outputting quantum states, making it a versatile subroutine for other quantum information tasks, rather than relying on classical post-processing as seen in quantum error mitigation or virtual state purification. This circuit architecture is particularly valuable as the transition to NISQ technologies necessitates effective error management and improved computational reliability for practical quantum applications.

“In most practical applications, having access to multiple copies of a resource state is entirely realistic,” the researchers write, suggesting QPA may achieve results comparable to quantum error correction while potentially using fewer resources. The protocol’s ability to improve performance with limited resources positions it as a promising strategy for advancing quantum information processing in the NISQ era.

Theoretical Foundation: d-Dimensional States & Symmetry

Previous quantum purity amplification (QPA) studies, beginning roughly two decades ago, largely focused on states possessing simple symmetries, single-qubit and coherent states, due to optimization challenges inherent in systems lacking such symmetry. Applying single-qubit protocols independently to multiple registers proved suboptimal for handling correlations within many-body systems, a limitation now addressed by a new theoretical framework. The theoretical underpinnings of this new protocol rely on d-dimensional states, known as qudits, providing a structured mathematical foundation for manipulating quantum information.

The protocol defines an (n,δ)-QPA protocol as a quantum channel that takes n copies of a noisy state and produces an output state with a fidelity exceeding a defined threshold, applicable to all depolarized d-dimensional pure states. Efficiency is achieved through a circuit architecture scaling polynomially in n and logarithmically in d, with each qudit implementable using a number of qubits equal to the ceiling of the base-2 logarithm of d. The algorithm employs a control register alongside a data register holding the states undergoing processing.

Schur Sampling, Correction, and Trace-Out Protocol Steps

The optimal quantum purity amplification (QPA) protocol relies on a defined sequence of operations beginning with Schur sampling, a process that categorizes input states based on their symmetry properties. This initial step projects states onto irreducible representations, effectively organizing them into sectors defined by differing symmetries, and establishes a block diagonal form key for subsequent processing. Unlike prior methods that discarded data from certain symmetry sectors, this protocol retains all information, processing each mixed symmetry sector to preserve potentially valuable quantum characteristics.

Following Schur sampling, a correction step relocates the most favorably symmetrized state to the final register, preparing the data for purification. This correction is applied conditionally, based on the outcome of the initial symmetry measurement, and is key to maximizing the fidelity of the final output. The protocol then concludes with a trace-out operation, discarding excess registers and isolating the purified output state.

An efficient algorithm, detailed in the supporting documentation, facilitates practical implementation of these three steps, using the Choi matrix for each outcome to ensure accurate processing. The algorithm’s design allows for the processing of each Schur sampling result through a correction stage followed by a partial trace, streamlining the purification process. This three-step process, combining Schur sampling, correction and trace-out, provides a pathway toward improving the performance of quantum information processing tasks, particularly for near-term devices.

Generalized Quantum Phase Estimation Enables Efficient QPA

This conversion ensures compatibility with standard qubit architectures, a critical step toward practical implementation of QPA techniques. The efficiency of this approach is further bolstered by an algorithm achieving gate complexity that scales polynomially with both the number of qudits and the logarithm of the dimension, a significant improvement over previous methods. Beyond algorithmic efficiency, the team introduced SWAPNET, a circuit designed specifically for near-term quantum devices, prioritizing low depth and gate efficiency.

Simulations, conducted in both digital and analog quantum environments, demonstrate the protocol’s resilience against realistic noise models, resulting in improved fidelity for tasks like Hamiltonian simulation and adiabatic evolution. These results align with the theoretical predictions underpinning the new QPA protocol, validating its effectiveness in mitigating the effects of noise. Experiments using superconducting quantum processors further substantiate these findings, confirming the protocol’s ability to enhance state preparation.

The researchers detail a structured approach to solving the optimization task inherent in QPA, formally defining the problem and presenting its solution in a manner that lends itself to operational interpretation. This work is divided into sections addressing the problem’s formalization, the solution itself, an efficient implementation, and finally, the results of numerical studies. The implementation’s runtime is defined as 4T_F + 2T_(CP) + T_(M-P), where T_F represents the time for a generalized quantum Fourier transform, T_(CP) the time for controlled permutations, and T_(M-P) the time for measurement and preparation.

SWAPNET Circuit Design for Near-Term Quantum Experiments

The newly developed SWAPNET circuit addresses a critical limitation of near-term quantum devices: the difficulty of implementing deep circuits. Designed to enable quantum purity amplification (QPA), SWAPNET bypasses complex gate compilation, offering a streamlined approach to improving quantum state fidelity. This circuit is particularly valuable as it allows for the conversion of QPA to systems requiring only ⌈ log(d)⌉ qubits, ensuring compatibility with existing hardware architectures.

The efficient implementation of SWAPNET relies on a gate complexity that scales polynomially with system size and the dimensionality of the quantum states being processed. Beyond its architectural advantages, SWAPNET’s performance has been validated through rigorous testing. The development of SWAPNET is presented alongside a formalization of the QPA optimization task, and a detailed solution that provides an operational interpretation of the process.

Superconducting Processor Validation of QPA Robustness

Experiments using IBM’s Marrakesh superconducting heavy-hex processor and the SamplerV2 primitive further substantiated these findings. The team focused on amplifying the state, measuring fidelity as the fraction of measurement outcomes matching this ideal state. To streamline the process, separate circuits were run for each iteration of SWAPNET, analyzing measurement outcomes that satisfied specific conditions to isolate the amplified signal. Initial simulations on a “FakeMarrakesh” processor, employing a static noise profile, generated 100,000 randomized circuit instances, each with two shots.

These simulations were then mirrored on the experimental Marrakesh processor, running 22,000 randomized instances with two shots each. Additional simulations explored higher qudit dimensions, k=4 and k=5, corresponding to d=2^4 and d=2^5, under conditions of global noise only. The team’s work demonstrates the practical utility of QPA.

QPA Applications: Hamiltonian Simulation & State Preparation

Quantum purity amplification (QPA) offers a direct route to improved quantum states, bypassing the need for classical post-processing common in error mitigation techniques or the expectation value estimations of virtual state purification. This direct output of amplified quantum states positions QPA as a valuable subroutine for a range of quantum information tasks, including quantum simulation, sensing and the preparation of states for quantum cryptography and machine learning applications.

The protocol’s utility extends to adiabatic state preparation, where simulations demonstrate fidelity gains even with increasing noise. Simulations exploring adiabatic state preparation with a one-dimensional Z₂ state reveal that QPA, when applied with multiple trials, consistently outperforms baseline methods. Specifically, the fidelity of state preparation improves as the number of trials increases, even under conditions of depolarization. Researchers tested this with a Hamiltonian tuned to prepare the state, optimizing the ramp profile to maximize input state fidelity for the QPA protocol.

The protocol’s performance was further validated through simulations mirroring the characteristics of real quantum hardware. These simulations incorporated noise modeled by local jump operators and coherence time limitations, providing a realistic assessment of QPA’s robustness. Results from both simulated and experimental processors demonstrate that QPA can achieve significant fidelity gains, even in the presence of substantial depolarization.

More like this

Quantum Research NewsMemoryless circuits halt entanglement, but retain nonclassical features

Quantum Research NewsQuantum systems show randomness from both the world and within

Quantum Error CorrectionFaultScope software pinpoints errors in quantum error correction circuits

Quantum Research NewsResearchers Map 281 Artworks Engaging Quantum Tech