Architectural Design of Scalable Quantum Computing Systems

Abstract

Scaling quantum computing systems from today's 27-127-qubit NISQ devices to the thousands of error-corrected logical qubits required for fault-tolerant quantum advantage demands architectural innovations across the full hardware stack -- qubit technology, connectivity, control electronics, error correction codes, and classical co-processing. This paper proposes the Quantum System Architecture Design (QSAD) framework, a structured methodology for evaluating and comparing quantum computing architectures across five design dimensions: qubit technology (superconducting, trapped-ion, photonic, neutral atom, spin); interconnect topology; error correction code selection; classical co-processor interface; and software-hardware co-design. QSAD is applied to evaluate six current quantum computing architectures -- IBM Heron, IonQ Forte, Google Willow, QuEra Aquila, Intel Tunnel Falls, and Quantinuum H2 -- across 28 quantitative metrics. IBM Heron and Quantinuum H2 score highest on QSAD composite (0.848 and 0.864 respectively). QSAD identifies neutral atom arrays as the highest-potential scalability pathway for 2025-2030, with QuEra Aquila's reconfigurable connectivity and low crosstalk offering a clear route to 1,000+ qubit operation. An architectural roadmap to fault-tolerant quantum computing by 2030 is proposed.