Calibration-Aware Transpilation for Variational Quantum Optimization

Ji, Yanjun; Brandhofer, Sebastian; Polian, Ilia

doi:10.1109/qce53715.2022.00040

“…The next crucial step is to identify high-quality qubits to execute them. To evaluate the qubit quality, we first consider only the circuit fidelity [15]. As illustrated in Fig.…”

Section: B Methodologymentioning

confidence: 99%

“…Mapping the algorithm's qubits to quantum computer's qubits (the qubit mapping problem) has been shown to be NP-hard. Various transpilation approaches [15][16][17][18][19][20][21][22] were designed. Furthermore, many heuristic [2,[23][24][25][26][27][28] and exact [2,23,[29][30][31][32][33] methods were developed to solve qubit mapping problem.…”

Section: B Noise-aware Transpilationmentioning

confidence: 99%

Optimizing QAOA on Bipotent Architectures

Ji¹,

Koenig²,

Polian³

2023

Preprint

0

View full text Add to dashboard Cite

Vigorous optimization of quantum gates has led to bipotent quantum architectures, where the optimized gates are available for some qubits but not for others. However, such gate-level improvements limit the application of user-side pulse-level optimizations, which have proven effective for quantum circuits with a high level of regularity, such as the ansatz circuit of the Quantum Approximate Optimization Algorithm (QAOA). In this paper, we investigate the trade-off between hardware-level and algorithm-level improvements on bipotent quantum architectures. Our results for various QAOA instances on two quantum computers offered by IBM indicate that the benefits of pulse-level optimizations currently outweigh the improvements due to vigorously optimized monolithic gates. Furthermore, our data indicate that the fidelity of circuit primitives is not always the best indicator for the overall algorithm performance; also their gate type and schedule duration should be taken into account. This effect is particularly pronounced for QAOA on dense portfolio optimization problems, since their transpilation requires many SWAP gates, for which efficient pulse-level optimization exists. Our findings provide practical guidance on optimal qubit selection on bipotent quantum architectures and suggest the need for improvements of those architectures, ultimately making pulse-level optimization available for all gate types.

show abstract

“…The qubit mapping problem has been shown to be NP-hard [10]. While a high-quality and stable solution calculated by exact methods improves the performance of algorithms on NISQ computers (e.g., [11]), the compilation time for finding the optimal solution grows exponentially with the circuit size. Many heuristic methods [7], [8], [12]- [15] have been developed to speed up this process.…”

Section: Introductionmentioning

confidence: 99%

Algorithm-Aware Qubit Mapping

Ji¹,

Polian²

2023

Preprint

1

0

View full text Add to dashboard Cite

<p>Algorithm-Aware (AlgAw) qubit mapping aims at directly providing the solutions to qubit mapping of algorithms with regular structures based on the algorithm's features. Although the exact method provides a high-quality solution, its compilation time grows exponentially with the circuit size. To improve its scalability, we propose the AlgAw qubit mapping. The main idea is to first determine the subcircuits in an algorithm to be mapped, then analyze the optimal solutions of small-scale subcircuits found by exact methods to obtain solutions of large-scale subcircuits, and finally reconstruct the entire circuit and assign parameters. Applying AlgAw to the Quantum Approximate Optimization Algorithm (QAOA) on linear and T-shaped subtopologies produces optimal and scalable solutions for arbitrary numbers of qubits and depths, which is critical to the algorithm's performance on Noisy Intermediate-Scale Quantum (NISQ) computers. Compared to Qiskit, Tket, and SWAP-Network, AlgAw produces the least number of CNOT gates and the lowest circuit depth. Furthermore, AlgAw takes only a few seconds to obtain a circuit with a hundred qubits that satisfies the connectivity constraints. The benchmarking results on Quantum Processing Units (QPUs) show that AlgAw qubit mapping yields higher values of approximation ratio than others. AlgAw can also be applied to other algorithms such as the Variational Quantum Eigensolver (VQE).</p>

show abstract

“…The qubit mapping problem has been shown to be NP-hard [10]. While a high-quality and stable solution calculated by exact methods improves the performance of algorithms on NISQ computers (e.g., [11]), the compilation time for finding the optimal solution grows exponentially with the circuit size. Many heuristic methods [7], [8], [12]- [15] have been developed to speed up this process.…”

Section: Introductionmentioning

confidence: 99%

Algorithm-Aware Qubit Mapping

Ji¹,

Polian²

2022

Preprint

View full text Add to dashboard Cite

<p>Algorithm-Aware (AlgAw) qubit mapping aims at directly providing the solutions to qubit mapping of algorithms with regular structures based on the algorithm's features. Although the exact method provides a high-quality solution, its compilation time grows exponentially with the circuit size. To improve its scalability, we propose the AlgAw qubit mapping. The main idea is to first determine the subcircuits in an algorithm to be mapped, then analyze the optimal solutions of small-scale subcircuits found by exact methods to obtain solutions of large-scale subcircuits, and finally reconstruct the entire circuit and assign parameters. Applying AlgAw to the Quantum Approximate Optimization Algorithm (QAOA) on linear and T-shaped subtopologies produces optimal and scalable solutions for arbitrary numbers of qubits and depths, which is critical to the algorithm's performance on Noisy Intermediate-Scale Quantum (NISQ) computers. Compared to Qiskit, Tket, and SWAP-Network, AlgAw produces the least number of CNOT gates and the lowest circuit depth. Furthermore, AlgAw takes only a few seconds to obtain a circuit with a hundred qubits that satisfies the connectivity constraints. The benchmarking results on Quantum Processing Units (QPUs) show that AlgAw qubit mapping yields higher values of approximation ratio than others. AlgAw can also be applied to other algorithms such as the Variational Quantum Eigensolver (VQE).</p>

show abstract

Calibration-Aware Transpilation for Variational Quantum Optimization

Cited by 6 publications

References 22 publications

Optimizing QAOA on Bipotent Architectures

Optimizing QAOA on Bipotent Architectures

Algorithm-Aware Qubit Mapping

Algorithm-Aware Qubit Mapping

Contact Info

Product

Resources

About