Quantum Computers Simplify Access Via Standard Software Connections

A new integration pathway for quantum computers into existing high-performance computing (HPC) centres has been demonstrated by Lukas Burgholzer and colleagues at Technical University of Munich. The pathway uses the Quantum Device Management Interface (QDMI), addressing the complexity arising from vendor-specific software chains. Implementing a QDMI layer with IQM superconducting systems and connecting it to Slurm job execution and Qiskit workflows sharply reduces the bespoke engineering required for each quantum backend. The resulting standardised software-hardware boundary provides reusable software components across different providers and deployment styles, enabling a faster transition from quantum pilots to production workflows.

Standardised interface dramatically reduces quantum software development overhead

The implementation reduces custom engineering effort by 75%, a figure previously unattainable due to the lack of standardised interfaces between quantum hardware and high-performance computing systems. This substantial reduction in effort is critical because the development of quantum algorithms and applications is already a complex undertaking, and the added burden of hardware-specific integration significantly hinders progress. Prior to this work, integrating a new quantum processor into an HPC environment often necessitated a complete rewrite of interfacing software, consuming valuable time and resources. This 75% reduction allows development teams to focus on algorithm design and optimisation, rather than low-level hardware communication. This threshold enables genuinely reusable software stacks, moving beyond isolated pilot projects and allowing scalable quantum integration within HPC centres. Previously, each quantum backend demanded entirely new software development, creating a significant operational burden, particularly for centres aiming to offer quantum computing as a service to a broad user base.

Central to this advancement is the Quantum Device Management Interface, or QDMI, which functions as a universal translator between quantum processors and conventional computing resources, decoupling software evolution from specific hardware characteristics. QDMI achieves this by defining a consistent set of application programming interfaces (APIs) for controlling and monitoring quantum hardware. These APIs abstract away the intricacies of the underlying quantum system, allowing software developers to interact with quantum computers in a standardised manner. This abstraction is analogous to the role of device drivers in conventional computing, shielding applications from the complexities of hardware implementation. A practical integration path for quantum hardware into high-performance computing environments is now available, utilising QDMI as a central component. A QDMI layer, backed by IQM superconducting systems, successfully connected to both Slurm-based job execution and Qiskit workflows, allowing for a functioning end-to-end quantum scientific computing workflow. The publicly available implementation, hosted on GitHub at https://github.com/iqm-finland/QDMI-on-IQM, streamlines operations by isolating vendor-specific characteristics within a device plugin. This plugin architecture allows for easy adaptation to different quantum hardware platforms without requiring modifications to the core QDMI layer. While this sharply reduces integration effort, challenges related to calibration updates, maintenance procedures, and the complexities of managing quantum resources at scale remain; a fully thorough HPCQC architecture remains a future goal. These challenges include ensuring the stability and accuracy of quantum computations over extended periods, as well as efficiently allocating and scheduling quantum resources among multiple users.

Facilitating quantum-classical integration with a superconducting processor interface

Attention is now focused on integrating quantum computers into existing high-performance computing centres, shifting the primary challenge from simply accessing the hardware to managing its complex interaction with conventional systems. The initial focus on hardware access, while necessary, proved insufficient for widespread adoption. The true potential of quantum computing lies in its ability to accelerate specific computational tasks when combined with the strengths of classical HPC systems. This requires seamless integration and efficient data transfer between the two paradigms. How readily QDMI will adapt to the diverse range of quantum architectures emerging, such as trapped ions or photonic systems, is a key consideration, as each possesses unique control and communication requirements. Trapped ion systems, for example, rely on different control signals and qubit connectivity compared to superconducting circuits. Photonic systems present further challenges related to single-photon detection and manipulation. Further development will need to address the challenges of supporting multiple quantum modalities and ensuring interoperability across different hardware platforms. This may involve developing a modular QDMI architecture that allows for the addition of new device plugins tailored to specific quantum technologies.

A standardised software layer between quantum hardware and existing high-performance computing (HPC) infrastructure reduces complexity for both users and operators, allowing centres to begin integrating quantum computers into workflows without custom engineering for each new device. This simplification is crucial for attracting a wider range of users to quantum computing, including those without extensive expertise in quantum hardware. By abstracting away the hardware details, QDMI allows researchers to focus on developing and running quantum algorithms without being bogged down in low-level implementation issues. This is a vital step towards practical quantum-classical computing, and the immediate value of this approach is not diminished by its current functionality with only IQM’s superconducting systems. The successful implementation demonstrates a pathway to reusable software components, a key requirement for scaling quantum computing resources. The ability to reuse software components across different quantum platforms will significantly reduce the cost and effort associated with deploying and maintaining quantum computing infrastructure.

Streamlining connections, rather than simply achieving hardware access, now defines successful integration of quantum computers into high-performance computing (HPC) centres. The paradigm has shifted from demonstrating the possibility of quantum-classical integration to demonstrating a practical and scalable solution. A standardised layer designed to decouple quantum hardware from user software provides a practical pathway for this integration. By implementing it with IQM superconducting systems and linking it to established HPC tools like Slurm, a system managing computer resource access, and the Qiskit software development kit, reusable software components are achievable. Slurm provides essential functionalities such as job scheduling, resource allocation, and monitoring, while Qiskit offers a comprehensive suite of tools for quantum algorithm development and simulation. The combination of these technologies, facilitated by QDMI, represents a significant step towards realising the full potential of quantum-classical computing.

The researchers successfully integrated quantum hardware into high-performance computing centres by establishing a standardised interface, the Quantum Device Management Interface (QDMI). This integration simplifies operations by decoupling software from specific hardware, meaning users can access quantum computers without bespoke engineering for each device. The implementation, tested with IQM superconducting systems alongside Slurm and Qiskit, demonstrates that standardising this software-hardware boundary is achievable today. This approach facilitates the development of reusable software components, which is essential for scaling quantum computing resources and attracting a broader user base.