Quantum Computers Need Unified Software to Unlock Their Full Potential

A thorough analysis of nine production quantum high-performance computing (QHPC) stacks has identified key design patterns and requirements across runtime, resource management, orchestration, and execution layers. Amir Shehata of Lawrence Berkeley National Laboratory, Oak Ridge National Laboratory, Technical University of Munich, Argonne National Laboratory, RIKEN, and colleagues found consistent needs for runtime abstraction, resource management, interconnect semantics, and observability. This analysis led to the proposal of the open quantum-HPC software ecosystem (openQSE) reference architecture, which unifies current practices and enables interoperability between different implementations. The architecture supports both noisy intermediate-scale quantum (NISQ) and future fault-tolerant quantum computing (FTQC) systems without altering application interfaces.

Interfaces and assumptions within diverse quantum high-performance computing stacks

Detailed architectural analysis underpinned this work, carefully dissecting nine production quantum high-performance computing (QHPC) stacks to reveal underlying commonalities. The process involved systematically reverse engineering each stack’s software layers and identifying how they handled tasks such as job submission, resource allocation, and data transfer; this is akin to carefully dismantling a complex clock to understand how each gear contributes to telling time. The analysis mapped interfaces and assumptions across the stacks, creating a convergence matrix to pinpoint areas of agreement and divergence, rather than simply cataloguing features.

AWS, IonQ, and Quantinuum systems were among the nine production quantum high-performance computing (QHPC) stacks analysed to identify common design patterns and emerging requirements. The focus of the analysis was on deployment models, application interaction, and software development kit (SDK) support, excluding detailed hardware specifications. This approach addressed fragmentation across the field and reduced maintenance costs by avoiding the need to repeatedly rebuild integration logic. A standardised approach to integration is needed, paving the way for more efficient resource utilisation and simplified development workflows.

A standardised openQSE architecture enables interoperability across diverse quantum computing

Designs from nine production quantum high-performance computing (QHPC) stacks have been unified into a single reference architecture, openQSE, overcoming the previous isolation of these systems. Custom integration was previously required for each of the nine analysed stacks when integrating quantum resources with existing high-performance computing infrastructure; now, a standardised blueprint enables interoperability without altering existing application interfaces. This advancement allows developers to write code once and deploy it across diverse quantum hardware, reducing maintenance costs and accelerating innovation in the field.

Quantum resources are increasingly integrated into high-performance computing (HPC) and cloud environments, but quantum high-performance computing (QHPC) software stacks often remain isolated proprietary solutions lacking common interfaces across runtime, resource management, orchestration, and execution layers. An analysis of nine production QHPC stacks identifies common design patterns and emerging requirements, covering deployment models, application interaction patterns, SDK support, and readiness for fault-tolerant operation. The survey reveals consistent needs in runtime abstraction, resource management, interconnect semantics, and observability, leading to the proposal of the open quantum-HPC software ecosystem (openQSE) reference architecture as a step toward unifying current practices. This architecture defines layer boundaries allowing interoperability while preserving deployment flexibility, and supports both noisy intermediate-scale quantum (NISQ) workloads and future fault-tolerant quantum computing (FTQC) systems without altering upper-layer application interfaces.

OpenQSE delivers a reference architecture for interoperable quantum and classical computing

Establishing a unified architecture for quantum high-performance computing (QHPC) promises to unlock greater potential from increasingly integrated quantum resources. This work, detailing the open quantum-HPC software ecosystem (openQSE), addresses the current fragmentation of proprietary stacks and the lack of standardised interfaces. The authors acknowledge a gap in fully implemented fair-share queuing features, similar to those found in established HPC systems like Slurm, suggesting that automated, equitable resource allocation remains a challenge and could limit the seamless integration of quantum tasks into existing workloads.

Despite current limitations in fully realised fair-share queuing, this work represents a key step forward for quantum high-performance computing. Establishing openQSE, a reference architecture defining interoperable layers for runtime, resource management, and application interaction, tackles a critical issue: the existing fragmentation of proprietary quantum software. This unified approach will support innovation and accelerate the integration of quantum tasks alongside conventional workloads, even without immediate perfect resource allocation.

Analysis of nine diverse production quantum high-performance computing stacks has yielded a unifying reference architecture, openQSE, designed to address the current lack of interoperability within the field. This blueprint establishes consistent layer boundaries for runtime, resource management, and application interaction, allowing different quantum systems to work together more effectively; for example, runtime abstraction simplifies how software communicates with varied quantum hardware. By identifying shared design patterns, scientists have created a foundation for a more cohesive ecosystem, potentially accelerating the development of hybrid quantum-classical applications and fostering a more collaborative research environment.

The research revealed a unifying reference architecture, openQSE, designed to improve interoperability between diverse quantum high-performance computing stacks. This architecture defines consistent layers for runtime, resource management, and application interaction, addressing the current fragmentation of proprietary software. Analysis of nine production systems highlighted consistent needs in these areas, and openQSE is structured to support both current noisy intermediate-scale quantum and future fault-tolerant systems. The authors propose this as a first step towards a more unified and collaborative quantum computing ecosystem.