en.Wedoany.com Reported - Qoro has launched a software development kit called Divi, designed to address the inefficiency of high-volume circuit generation in variational quantum algorithms. Compared to existing frameworks such as PennyLane, Qiskit, and Cirq, Divi treats batch generation as a composable pipeline, achieving up to a 148x improvement in pipeline execution speed. The company notes that performing useful work on NISQ hardware typically involves running thousands of nearly identical circuit variants rather than a single circuit. Divi outputs standard OpenQASM, maintaining vendor neutrality, offering researchers a new option to optimize cloud costs and accelerate result acquisition.
An optimizer step in variational quantum algorithms rapidly generates a large number of circuits to be executed, expressed as N_groups × K × T × P. While current quantum computing frameworks can construct individual circuits, they face computational bottlenecks when preparing and submitting these large-scale circuit batches. Xanadu's PennyLane introduced composable circuit transformations, IBM's Qiskit rewrote the circuit core in Rust (qiskit._accelerate), Google's Cirq treats parameter sweeps as first-class execution primitives, and Mitiq developed frontend-agnostic error mitigation techniques. However, these advances remain fragmented. Qoro's Divi aims to integrate these existing technologies into a unified pipeline, with the core principle of treating the entire batch as a single, composable pipeline operating on templated circuits. Qiskit's rewrite of its circuit core in Rust and Cirq's parameter sweep execution also informed Divi's development.
Divi's core innovation lies in deferring parameter binding to the final serialization stage. While other frameworks materialize parameter sweeps into multiple circuit copies before submission, Divi's pipeline composes ordered stages such as grouping, folding, twirling, and basis rotation, operating on parameter-free templates. The circuit body is serialized only once, and parameter sets are then efficiently substituted, ensuring that expensive, parameter-independent computations are performed only once and shared across all circuit variants. Qoro reports that Divi achieves up to a 148x improvement in pipeline execution speed compared to benchmarks with PennyLane, Qiskit, and Cirq. The system measures performance in terms of both in-memory batch generation and fully serialized batch output.
Divi maintains vendor neutrality by outputting standard OpenQASM, avoiding proprietary serialization schemes found in other frameworks, such as Qiskit's QPY or Google's custom protobuf. Qoro emphasizes that each framework provides excellent components but leaves the task of assembling batches to the user. The Divi open-source package is available via pip install qoro-divi. Qoro also offers a commercial platform, Qoro Solo, for users to test the pipeline on real-scale quantum algorithms. The team believes that for useful work on NISQ hardware, the speed bottleneck often lies in reconstructing and re-serializing circuit batches, rather than executing the circuits themselves.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
