Palsberg and Colleagues Presents Language Classification Framework for Quantum Programming

Quan Do and colleagues at the University of California, Los Angeles have surveyed ten popular, executable quantum programming languages using a new classification framework encompassing programming model, expressiveness, and safety. The survey delivers implementations of fifty benchmark programs, including Shor’s algorithm for integer factoring and Hamiltonian simulations, and an automated testing harness to measure program performance and correctness. The University of California, Los Angeles team assessed ten languages using a new framework focused on how programs are built, their capabilities, and security features.

They implemented fifty benchmark programs, including key algorithms like Shor’s for factoring and simulations of quantum systems, alongside automated testing to verify performance and accuracy. This thorough survey reveals the strengths and weaknesses of each language through practical application. Quan Do and colleagues at the University of California, Los Angeles have undertaken a detailed comparison of ten popular quantum programming languages, moving beyond conceptual discussions to assess concrete implementations.

The team developed a new classification framework evaluating languages based on their programming model, a set of tools dictating how a programmer interacts with the quantum computer, their expressiveness, and safety features. They implemented fifty benchmark programs, including vital algorithms such as Shor’s for integer factoring and Hamiltonian simulation, a technique for modelling quantum systems. This detailed survey reveals each language’s strengths and weaknesses through practical application.

Benchmarking quantum languages via standardised algorithms and Hamiltonian simulations

A standardised benchmarking technique was employed, implementing fifty identical programs across ten languages to facilitate direct comparison. The evaluation extended beyond mere program executability, focusing on efficiency and complexity, comparable to assessing painters by requesting they all recreate the same field. Specifically, the programs included Shor’s algorithm, a complex calculation for factoring large numbers, and Hamiltonian simulations, a technique for modelling quantum systems akin to creating a virtual wind tunnel for aircraft design.

Fifty identical programs, including Shor’s algorithm and Hamiltonian simulations, were used to evaluate ten languages. These simulations utilise Hamiltonians, such as the transverse-field Ising model and the Heisenberg isotropic chain, to represent quantum system evolution, chosen for their prominence in existing benchmark suites. This standardised approach allowed for direct comparison of language efficiency and complexity, exceeding evaluations based solely on whether programs could run.

Thorough benchmarking reveals performance across ten languages

A a five-fold increase in the scale of executable benchmark programs across multiple quantum programming languages has been achieved, expanding from previous limited surveys to a thorough suite of fifty programs. This leap enables meaningful comparisons of language performance and capabilities, previously impossible due to a lack of standardised, executable tests; earlier work relied heavily on conceptual analysis rather than practical implementation. The University of California team developed a new classification framework to assess ten languages, focusing on programming model, expressiveness, and safety, and used this to implement algorithms like Shor’s for integer factoring and Hamiltonian simulations.

The University of California researchers expanded their benchmark suite to fifty programs, revealing that seven of the ten languages tested successfully implemented Shor’s algorithm, a computationally intensive task, demonstrating a strong level of practical capability. All ten languages could also express Hamiltonian simulations, specifically the transverse-field Ising model and the Heisenberg isotropic chain, utilising both Trotterization and Linear Combination of Unitaries methods to approximate quantum system evolution. Detailed analysis showed program sizes varied considerably, ranging from approximately 150 lines of code for the most concise implementations to over 500 lines for the most verbose, highlighting differences in language verbosity and abstraction levels. Despite these advances, the team acknowledges that these fifty programs represent relatively simple workloads; scaling these benchmarks to realistically complex problems, mirroring those anticipated to benefit from quantum advantage, remains a substantial hurdle.

Benchmarking quantum programming languages reveals limitations and commonalities

Unlocking the potential of quantum computers requires more than just faster hardware, as scientists are striving to achieve. Programmers need accessible tools, and the proliferation of quantum programming languages, ten languages were surveyed in this study, creates a fragmented field. While the University of California team successfully implemented key algorithms across these languages, the limited scope of the benchmarks raises an important consideration: can these results accurately predict performance on the complex, real-world problems where quantum computers are expected to excel.

Even acknowledging the limited scope of the University of California team’s benchmarks, this work represents a vital step forward for quantum software development. Identifying common ground and challenges across these ten languages, including PennyLane and Qiskit, tools designed to make quantum computing more accessible, is key for building a robust ecosystem. A standardised approach will ultimately accelerate progress beyond theoretical demonstrations and towards practical applications in fields like materials science and drug discovery.

The University of California team established a new framework for evaluating quantum programming languages, moving beyond simple assessments of functionality to consider program construction, expressive power, and inherent safety features. This systematic approach enabled a comparative survey of ten popular languages, revealing a diverse range of design choices and trade-offs in their implementation; the languages included PennyLane and Qiskit, tools aimed at broadening access to quantum computation. Implementing fifty benchmark programs, including Shor’s algorithm, a calculation for factoring numbers, and Hamiltonian simulations, provided the researchers with concrete data for assessing each language’s strengths and weaknesses.

The research successfully classified and compared ten languages, including PennyLane and Qiskit, using a new evaluation framework. This work highlights the current challenges in quantum software development as the field moves towards more complex programs. By systematically assessing program construction, expressive power, and safety features across these languages, researchers identified commonalities and areas for improvement. The team implemented fifty benchmark programs, such as Shor’s algorithm and Hamiltonian simulations, to provide a concrete basis for comparison and inform future language design.