# Optimizing QAOA on Bipotent Architectures

**Authors:** Yanjun Ji, Kathrin F. Koenig, Ilia Polian  
**arXiv:** [2303.13109](https://arxiv.org/abs/2303.13109)  
**Date:** 2023-03-23  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2303.13109v1/

## Key points

- Quantum computing has made progress in optimizing quantum gates, resulting in bipotent quantum architectures where optimized gates are available for some qubits but not for others.
- This limits the application of user-side pulse-level optimizations that have proven effective for regular quantum circuits such as the ansatz circuit of the Quantum Approximate Optimization Algorithm (QAOA).
- A study was conducted to investigate the benefits of pulse-level optimizations versus vigorously optimized monolithic gates on bipotent quantum architectures.
- Results showed that pulse-level optimizations currently outweighed the improvements due to vigorously optimized monolithic gates for various QAOA instances on two IBM quantum computers.
- Circuit primitives' fidelity is not always the best indicator of overall algorithm performance; gate type and schedule duration should also be taken into account.
- The findings provide practical guidance on optimal qubit selection on bipotent quantum architectures and suggest the need for improvements to make pulse-level optimization available for all gate types.
- Balancing hardware and algorithmic improvements is important in optimizing quantum circuits effectively to maximize performance.

## Summary

In recent years, quantum computing has seen significant progress in the optimization of quantum gates, resulting in bipotent quantum architectures where optimized gates are available for some qubits but not for others. This gate-level optimization limits the application of user-side pulse-level optimizations that have proven effective for regular quantum circuits such as the ansatz circuit of the Quantum Approximate Optimization Algorithm (QAOA). To address this trade-off between hardware and algorithm-level improvements on bipotent quantum architectures, Yanjun Ji, Kathrin F. Koenig, and Ilia Polian conducted a study investigating the benefits of pulse-level optimizations versus vigorously optimized monolithic gates. Their results showed that pulse-level optimizations currently outweighed the improvements due to vigorously optimized monolithic gates for various QAOA instances on two IBM quantum computers. Additionally, their data indicated that circuit primitives' fidelity is not always the best indicator of overall algorithm performance; gate type and schedule duration should also be taken into account. This effect was particularly pronounced for QAOA on dense portfolio optimization problems since their transpilation requires many SWAP gates for which efficient pulse-level optimization exists. The findings provide practical guidance on optimal qubit selection on bipotent quantum architectures and suggest the need for improvements to make pulse-level optimization available for all gate types. Overall, this study highlights the importance of balancing hardware and algorithmic improvements in optimizing quantum circuits effectively to maximize performance.