Qanela: A framework proposal for designing portable quantum assembly languages across different quantum computing platforms

Abstract

The diversity of quantum hardware platforms creates significant algorithm portability challenges, with each platform employing distinct instruction sets, gate implementations, and physical constraints that hinder software reusability. This paper presents a framework for designing portable quantum assembly languages enabling seamless cross-platform execution. Through systematic profiling of Shor’s, Grover’s, and QAOA algorithms on IBM Qiskit, Microsoft Q#, and Google Cirq — 3,216 executions of identical, reference-validated logical circuits — we identify key portability challenges and derive empirically grounded design criteria from execution-pattern analysis. With matched workloads, cross-platform circuit-depth divergence is 1.3–1.6× where no gate decomposition is required and rises to 94.8× under multi-controlled decomposition pressure, of which 3.36× is genuine same-level synthesis divergence and the remainder an abstraction-level measurement gap; optimization semantics, topology expressibility, and quantum–classical execution contracts diverge analogously. The framework balances hardware abstraction with performance optimization through declared abstraction levels, topology-aware compilation, optimization profiles with defined semantics, and a universal gate set with platform-specific decomposition. Prototype validation demonstrates 100% automated compilation of unmodified single-source programs across 30 platform configurations, with all 90 executions passing calibrated distributional acceptance gates and exact statevector fidelity of 1.000000 against the language’s reference semantics, providing feasibility evidence for vendor-independent quantum software development.