Researchers Unify Quantum Provenance across Three Platforms

Researchers from the University of Jyväskylä, working with collaborators at MIT and Università della Svizzera italiana, have developed a framework to standardize provenance tracking across quantum computing platforms. Their study identifies a major obstacle to reliable quantum software development: the lack of consistent methods for documenting how quantum programs are compiled, executed, and processed on different hardware. After analysing 15 quantum platforms spanning five hardware technologies, the team found that provenance information is often fragmented, incomplete, and difficult to compare across providers.

To address this challenge, the researchers propose an evidence-aware OpenAPI 3.1 provenance contract together with a provider-adapter architecture that creates a unified representation of provenance data while preserving provider-specific information. A reference prototype developed at QMill demonstrates the approach using Amazon Braket Quantum and IonQ platforms, with all generated records successfully validating against a single common specification.

Provenance data records the complete lifecycle of a quantum computation, including the quantum program, compilation steps, execution details, hardware characteristics, software environment, and resulting outputs. Such information is essential for reproducing experiments, verifying results, debugging software, and comparing the performance of different quantum systems. However, today’s quantum providers expose this information through diverse software development kits, application programming interfaces, and proprietary data models, making consistent analysis difficult.

The researchers evaluated publicly available provenance information from 15 quantum platforms representing superconducting, trapped-ion, neutral-atom, photonic, and quantum-annealing technologies. Their analysis revealed significant inconsistencies across providers, with compilation provenance proving to be the least complete category. Because the compilation process translates quantum circuits into hardware-specific instructions that can substantially affect computational outcomes, missing compilation data presents a major challenge for validating experimental results.

The proposed OpenAPI 3.1 contract establishes a machine-readable standard for describing provenance information, enabling automated validation while allowing hardware providers to retain implementation-specific details. Rather than requiring providers to adopt identical internal data structures, the provider-adapter architecture translates each platform’s existing provenance records into a common format. This design simplifies integration across multiple hardware platforms while remaining flexible enough to accommodate future quantum technologies.

The QMill reference implementation demonstrates that provenance records from Amazon Braket Quantum and IonQ can be successfully converted into the unified schema without losing provider-specific metadata. The framework also supports incomplete provenance records by explicitly representing missing information instead of discarding partially documented computations, allowing researchers to work with currently available data while encouraging more comprehensive reporting.

Beyond improving reproducibility, the framework provides a foundation for more reliable benchmarking, hardware validation, and software development as quantum computing continues to mature. A standardized provenance system would enable researchers to compare results generated on different hardware platforms with greater confidence while simplifying collaboration across institutions and software ecosystems.

The authors note that broader adoption will depend on quantum hardware providers consistently exposing richer provenance information through their platforms. As quantum systems become increasingly sophisticated, standardized provenance tracking could become an essential component of scientific transparency, helping transform quantum computing from isolated demonstrations into a reproducible and dependable computational technology.