InferQ: New Benchmark for Quantum Circuit Simulation via SQL
InferQ has been developed by researchers as a benchmark focused on databases to assess quantum circuit simulation through relational database management systems (RDBMSs). This method translates quantum circuit simulations into SQL workloads, mainly involving join-and-aggregate tensor contractions. InferQ constructs general, compositional circuits by combining subcircuits from predefined templates, producing each simulation task as an SQL workload suitable for RDBMS. It captures various features of circuits and queries—static, graph, SQL, and dynamic—for effective workload analysis. The benchmark features an extensive dataset of 202,975 circuits, accessible online with a web viewer for search and filtering. This initiative overcomes the constraints of earlier studies that concentrated on a limited array of structured circuits, enhancing support for comprehensive database research in quantum computing simulation.
Key facts
- InferQ is a database-oriented benchmark for quantum circuit simulation.
- It compiles quantum circuit simulation into SQL workloads (join-and-aggregate tensor contractions).
- InferQ generates general, compositional circuits from circuit templates.
- It emits each simulation task as an RDBMS-ready SQL workload.
- It extracts circuit and query features: static, graph, SQL, and dynamic.
- The dataset includes 202,975 circuits.
- A web-based viewer supports searching and filtering of the dataset.
- The work is published on arXiv with ID 2607.29134.
Entities
—