Quantum Compilers’ Retargetability Assessed: New Metric Analyses Key Aspects

Researchers are tackling the crucial challenge of ensuring quantum software can run seamlessly across diverse and rapidly evolving quantum hardware. Luke Southall, Joshua Ammermann, and Rinor Kelmendi, from the Karlsruhe Institute of Technology, alongside Domenik Eichhorn and Ina Schaefer, present a novel metric to rigorously assess the ‘retargetability’ of leading quantum compilers , essentially, how easily programs can be adapted for different machines. This work is significant because the proliferation of quantum computing approaches demands software independence, and their study, comparing Tket, Qiskit, and ProjectQ, reveals that Tket and Qiskit currently offer superior retargetability compared to ProjectQ. These findings will directly inform quantum software developers’ compiler choices and guide future improvements in this vital area of quantum software engineering.

These findings will directly inform quantum software developers’ compiler choices and guide future improvements in this vital area of quantum software engineering.

Quantum compiler adaptability across diverse hardware

Currently, the NISQ-era is characterised by a diverse landscape of quantum hardware manufacturers, each employing unique approaches and architectures, which complicates the task of creating software applicable across different platforms. This research responds to this problem by introducing a robust methodology for evaluating how effectively quantum compilers can adapt to these varied hardware constraints, akin to the challenges faced in classical computing when compiling for different processor types like ARM or x86. The team achieved a significant breakthrough by not only defining key dimensions of retargetability but also designing a user study to empirically measure this crucial characteristic in leading quantum compilers. The study meticulously analysed three prominent compilers, Tket, Qiskit, and ProjectQ, utilising a newly developed retargetability metric and a carefully designed user study involving six participants.
The research establishes a clear methodology for quantifying retargetability, identifying important dimensions and translating them into a measurable metric. The work opens avenues for future research focused on standardising the quantum computing stack and enhancing compilation techniques to facilitate broader adoption of retargetable compilers, ultimately accelerating progress in quantum software development. This breakthrough reveals the importance of compiler retargetability in the context of heterogeneous quantum hardware, where qubit connectivity, environmental requirements, and coherence times vary significantly between platforms. The study unveils that a compiler’s ability to generate valid instructions for diverse hardware architectures is paramount, and the developed metric provides a valuable tool for assessing this capability. Ultimately, this research contributes to building a future-proof quantum software infrastructure capable of harnessing the full potential of diverse quantum computing technologies.

Compiler Retargetability Evaluation via User Study reveals key

The research team recognised the increasing heterogeneity of quantum hardware and the consequent need for compilers capable of adapting to diverse architectures. This work pioneers a systematic approach to evaluating a compiler’s ability to generate valid quantum circuits for different hardware backends. Initially, researchers identified key dimensions of retargetability, forming the basis of their bespoke metric. Participants then evaluated these compilers based on the identified retargetability aspects, providing subjective ratings that were subsequently combined to generate a comparative score for each compiler.
The experimental setup demanded that each participant independently construct and assess backends, minimising bias and ensuring a robust evaluation. The study employed a rigorous data collection procedure, capturing participant ratings across multiple facets of retargetability. This data was then aggregated and analysed to produce a retargetability score for each compiler, allowing for direct comparison. The team harnessed the power of user expertise to evaluate the compilers, recognising that automated metrics alone may not fully capture the nuances of retargetability. This innovative approach moves beyond simple performance benchmarks, focusing instead on the compiler’s adaptability and ease of backend implementation. The team’s work establishes a foundation for standardised retargetability assessments, crucial for the advancement of quantum software engineering.

Tket Excels in Quantum Compiler Retargetability

Experiments revealed that Tket demonstrated the highest level of retargetability, indicating its superior ability to adapt to different quantum hardware architectures. Data shows Qiskit closely followed Tket in performance, while ProjectQ lagged behind in its capacity for cross-platform compatibility. The core of this work lies in a newly designed retargetability metric, built upon identifying crucial dimensions of compiler adaptability and applying them to a user study. Measurements confirm that Tket achieved the highest retargetability score, signifying its effectiveness in generating valid circuits for diverse hardware backends.

Qiskit received a comparable, though slightly lower, score, demonstrating strong retargetability capabilities, while ProjectQ’s score was notably lower, highlighting areas for potential improvement. Results demonstrate a clear distinction in the compilers’ abilities to handle the heterogeneity of quantum hardware, a significant obstacle in quantum software development. The study meticulously evaluated each compiler’s performance, providing a comparative analysis based on the implemented metric and user feedback. This breakthrough delivers a practical tool for navigating the complex landscape of quantum computing hardware.

Scientists recorded that the ability to compile software for multiple platforms is akin to the challenges faced in classical computing with architectures like ARM and x86, but is complicated by the lack of standardized architectures in the quantum realm. The research highlights that unlike the ubiquitous x86 and ARM architectures, quantum hardware varies significantly in environmental requirements, qubit connectivity, scalability, and coherence times. The findings underscore the importance of prioritizing retargetability in the development of quantum compilers to accelerate progress in the field.

Scientists Conclusion

This research involved a detailed study of three prominent compilers, Tket, Qiskit, and ProjectQ, evaluating them across five key dimensions: compilation strategy flexibility, standardization compliance, community and ecosystem integration, device-agnostic compiler architecture, documentation, and API quality. The established metric provides a structured approach to evaluating compiler retargetability, aiding informed decision-making and potentially accelerating the development of portable quantum software. The authors acknowledge limitations stemming from resource constraints, which restricted the number of compilers evaluated in the study. Future research could expand upon this work by analysing additional compilers, such as Weaver, and conducting further studies to generalise the findings and refine the assessment methodology.