PhysQ'26
Physics-Driven Approaches to Quantum Computing
From Device Physics to Simulation and Co-Design with High-Performance Computing
From Device Physics to Simulation and Co-Design with High-Performance Computing
July 6, Belfast, Northern Ireland, United Kingdom
Submission Deadline - April 15 SUBMISSIONS ARE OPEN
Notifications - May 1
Camera Ready Papers - TBD
Workshop - July 6
Quantum computing is often presented in terms of abstract qubits, circuits, and algorithms. Yet, real devices are strongly physical objects: many-body systems with specific Hamiltonians, open-system dynamics, control constraints, and thermodynamic costs. At the same time, the design, simulation, and operation of such devices increasingly depend on advanced classical computation, ranging from device-level electromagnetic solvers and multiscale modeling to hybrid quantum-HPC workflows and physics-informed quantum machine learning.
PhysQ focuses on physics-driven approaches to quantum computing: analog and continuous-variable (CV) models, open quantum systems and noise engineering, device-level modeling and simulation, thermodynamics and energetic considerations, and end-to-end workflows that connect experiments with advanced computing infrastructures. It also welcomes quantum machine learning (QML) work that is explicitly informed by device physics, noise, or hybrid HPC/quantum workflows. The goal is to treat physics itself as a primary design and a nalysis tool for quantum computation, while highlightinghow clusters, accelerators, and supercomputers enable these investigations and their translation into usable architectures, software, and QML pipelines.
We invite contributions that use physical models, open quantum systems, control, thermodynamics, advanced computing, or physics-aware QML to design, analyze, or simulate quantum computing platforms. Work that leverages substantial computational resources (e.g., clusters, GPU systems, supercomputers) is especially welcome but not required.
Analog quantum computing and Hamiltonian engineering for computation and optimization.
Continuous-variable (CV) and bosonic-mode approaches: physical models, encodings, and simulation techniques.
Comparisons of analog / CV implementations with digital / circuit implementations for the same computational task.
Hybrid schemes mixing gate-based, analog, and CV components in a single physical model.
Quantum thermodynamics and energy/entropy costs of computation, control and measurement.
Trade-offs between energy, speed, fidelity and cooling requirements in realistic devices.
Physics-inspired limits on clock speed, connectivity or error rates, and their implications for architecture and algorithm design.
Energy and cost models for hybrid quantum–classical or quantum–HPC workflows.
End-to-end workflows that couple experiments on quantum hardware with large-scale simulation or data analysis on advanced computing systems.
Data management, pipelines and analysis for large volumes of experimental or simulation data (e.g., calibration logs, tomography data, control traces).
Co-scheduling or coordinating quantum experiments with classical simulations (e.g., digital twins of devices, predictive modeling).
Experiences from computing centers or labs deploying advanced computing infrastructure in support of physics-based quantum hardware and workflows.
Numerical simulation of specific hardware platforms (superconducting, ion-trap,neutral-atom, photonic, spin qubits) using clusters, GPUs or supercomputers.
Multi-scale modeling: from device-level physics up to effective qubit-level models used in software stacks.
Physics-driven co-design loops that use large-scale simulation to optimize architecture, layout, control pulses or error-correction schemes.
Workflows and toolchains for systematic comparison of device designs, including how they run on advanced computing infrastructure.
Open-system models of realistic hardware (non-Markovian noise, cross-talk, leakage, drift).
Reservoir engineering and dissipation as computational resources rather than only errors.
Physics-based error-mitigation and error-correction strategies informed by microscopic noise models.
Large-scale numerical studies of decoherence, error processes and control strategies.
QML models and architectures explicitly informed by device physics, analog/CV dynamics, or realistic noise and control constraints.
Hybrid QML workflows that combine quantum devices, classical/HPC resources and physics-based simulators.
Benchmarking and evaluation of QML approaches using physics-motivated datasets or tasks (e.g., materials, many-body systems, HEP), including comparisons to strong classical baselines.
Training, validation, and uncertainty quantification for QML models using large-scale simulation or experimental data.
Position or methodology papers on when and how physics-based QML is expected to provide advantages over classical ML.
Stefano Markidis (KTH Royal Institute of Technology, Sweden)
Salvatore Mandrà (Google Quantum AI, USA)
Stefano Mensa (NVIDIA)
Oleksandr Kyriienko (University of Sheffield)
Erik M. Åsgrim (KTH Royal Institute of Technology, Sweden)