Periodic Reporting for period 2 - RoCCQeT (Real-World Commercial Coherent Quantum Annealing Technology)
Berichtszeitraum: 2024-03-01 bis 2025-08-31
Many of the world’s most valuable innovations start with accurate simulation—testing designs, materials, or molecules before building them. Yet some systems are so complex that today’s computers struggle to model them in a useful time. This is especially true in chemistry and materials science, where understanding electronic behaviour can accelerate cleaner energy technologies, better batteries, or new medicines.
Quantum computing offers a path to simulate such systems more efficiently by exploiting quantum physics itself. While fully error-corrected “digital” (i.e. gate-based) quantum computers remain a long-term goal, analogue approaches—tailored quantum devices that natively mimic specific models—can deliver earlier value with fewer qubits.
RoCCQeT aims to prototype analogue quantum processing units (AQPUs) based on superconducting circuits and to provide the software and algorithms needed to use them on real problems.
Our technical objectives are to:
- Develop and validate coherent, scalable hardware (fluxonium-based superconducting processors with flip-chip integration).
- Build the full software stack—control, cloud access, and developer tools—to run problems end-to-end.
- Co-design algorithms and embeddings with the hardware to target quantum chemistry and other industrially relevant use cases.
- Demonstrate practical workflows with early adopters, paving the way to pre-commercial, cloud-accessible quantum services.
By focusing first on chemistry simulations (e.g. small to medium molecules), the project addresses European priorities in energy, sustainability, and health, while strengthening technological autonomy in quantum technologies.
Quantum computing offers a path to simulate such systems more efficiently by exploiting quantum physics itself. While fully error-corrected “digital” (i.e. gate-based) quantum computers remain a long-term goal, analogue approaches—tailored quantum devices that natively mimic specific models—can deliver earlier value with fewer qubits.
RoCCQeT aims to prototype analogue quantum processing units (AQPUs) based on superconducting circuits and to provide the software and algorithms needed to use them on real problems.
Our technical objectives are to:
- Develop and validate coherent, scalable hardware (fluxonium-based superconducting processors with flip-chip integration).
- Build the full software stack—control, cloud access, and developer tools—to run problems end-to-end.
- Co-design algorithms and embeddings with the hardware to target quantum chemistry and other industrially relevant use cases.
- Demonstrate practical workflows with early adopters, paving the way to pre-commercial, cloud-accessible quantum services.
By focusing first on chemistry simulations (e.g. small to medium molecules), the project addresses European priorities in energy, sustainability, and health, while strengthening technological autonomy in quantum technologies.
1) Hardware. The consortium refined its hardware roadmap by pivoting from earlier flux-qubit designs to fluxonium qubits integrated via flip-chip packaging. This improves noise resilience and routing for larger devices. We fabricated the first generation of multi-qubit processors (up to 15 qubits), and validated single-/two-qubit building blocks with 8–10 μs coherence in initial measurements. Calibration methods (flux spectroscopy, two-tone fits, crosstalk compensation) were established and transferred to the new platform.
2) Control and cloud access. We delivered a modular local control stack with automated calibration pipelines, and a cloud pathway comprising a Public API for external users and a Lab API for secure on-premise execution. A web portal (beta → v2) and operational dashboards support job submission, monitoring, and system health, enabling multi-user access and transparent operation.
3) Algorithms and co-design. We released QiliSDK, an open developer toolkit with an Analog Transpiler that converts abstract time-dependent Hamiltonians into device-ready control signals. The team explored custom annealing schedules and introduced a triangular embedding method that maps fully connected problems onto hardware with local connectivity—reducing overhead and informing hardware design choices.
4) Pilots. In chemistry, we validated workflows using external quantum platforms while in-house devices proceed through cryogenic validation. Benchmarks such as H2/H₄ ground-state estimations were used to compare annealing-based and digital approaches, informing schedule design and tooling. Early adopters exercised the APIs, portal, and SDK, feeding back requirements that improved usability and reliability.
Together, these results establish an end-to-end path from problem definition to execution on analogue hardware, with a clear upgrade route as devices scale.
2) Control and cloud access. We delivered a modular local control stack with automated calibration pipelines, and a cloud pathway comprising a Public API for external users and a Lab API for secure on-premise execution. A web portal (beta → v2) and operational dashboards support job submission, monitoring, and system health, enabling multi-user access and transparent operation.
3) Algorithms and co-design. We released QiliSDK, an open developer toolkit with an Analog Transpiler that converts abstract time-dependent Hamiltonians into device-ready control signals. The team explored custom annealing schedules and introduced a triangular embedding method that maps fully connected problems onto hardware with local connectivity—reducing overhead and informing hardware design choices.
4) Pilots. In chemistry, we validated workflows using external quantum platforms while in-house devices proceed through cryogenic validation. Benchmarks such as H2/H₄ ground-state estimations were used to compare annealing-based and digital approaches, informing schedule design and tooling. Early adopters exercised the APIs, portal, and SDK, feeding back requirements that improved usability and reliability.
Together, these results establish an end-to-end path from problem definition to execution on analogue hardware, with a clear upgrade route as devices scale.
- Hardware architecture: The combination of fluxonium qubits and flip-chip integration provides a scalable, low-crosstalk platform for coherent analogue computation, addressing wiring density and isolation challenges that appear beyond ~10 qubits.
- Executable co-design: QiliSDK enables reproducible, portable workflows: users define Hamiltonians and schedules once, then run them on simulators or hardware with the same interface. The Analog Transpiler closes the loop between theory and experiment.
- Efficient embeddings and schedules: The triangular embedding scheme offers a principled way to deploy fully connected problems on locally connected devices; schedule engineering (time-dependent biases and ramps) improves robustness on near-term processors.
- Pilot-driven readiness: Cloud APIs, a developer SDK, and a portal shorten the time from research to pilot deployment and prepare for future Quantum-as-a-Service offerings.
Key needs for uptake: Continued maturation of fabrication to increase coherence and yield; demonstration on 6–10+ qubit subsystems; standardised interfaces and benchmarks; initial market pilots in chemistry/materials; and sustained investment to scale cloud access and user support. These steps will accelerate adoption and de-risk the transition from prototypes to dependable services.
- Executable co-design: QiliSDK enables reproducible, portable workflows: users define Hamiltonians and schedules once, then run them on simulators or hardware with the same interface. The Analog Transpiler closes the loop between theory and experiment.
- Efficient embeddings and schedules: The triangular embedding scheme offers a principled way to deploy fully connected problems on locally connected devices; schedule engineering (time-dependent biases and ramps) improves robustness on near-term processors.
- Pilot-driven readiness: Cloud APIs, a developer SDK, and a portal shorten the time from research to pilot deployment and prepare for future Quantum-as-a-Service offerings.
Key needs for uptake: Continued maturation of fabrication to increase coherence and yield; demonstration on 6–10+ qubit subsystems; standardised interfaces and benchmarks; initial market pilots in chemistry/materials; and sustained investment to scale cloud access and user support. These steps will accelerate adoption and de-risk the transition from prototypes to dependable services.