Quantum Simulations of Semiconductor Spectroscopy

This repository provides a framework for simulating spectroscopic experiments of semiconductor materials using digital quantum computers. The framework combines Qiskit quantum simulations with MPI-enabled parallel execution and efficient classical solvers, making it well suited for high-performance computing (HPC) environments.

• Python 3.9+
• Qiskit
• All libraries listed in requirements.txt

Before running the application, ensure that all required Python packages are installed.

1. Clone the repository:

git clone https://github.com/freude/sbe_qc.git
cd sbe_qc

2. Create and activate a virtual environment (recommended):

python -m venv venv
source venv/bin/activate        # Linux / macOS
venv\Scripts\activate           # Windows

3. Install the dependencies:

pip install -r requirements.txt

4. Optional: MPI Support to use parallel execution

sudo apt install openmpi-bin openmpi-common libopenmpi-dev
pip install mpi4py

To start the demo:

python run_engine.py

This launches the main application.

This file defines all runtime parameters for the simulation, including system size, band-structure properties, dynamical evolution, external pulse characteristics, backend execution options, and noise modeling. The configuration is organized into logical sections to simplify experimentation and reproducibility.

General system-level parameters.

• N_sites
  Number of lattice sites (or discretization points) used to represent the system.

• k_max
  Maximum wave vector value defining the extent of reciprocal-space sampling.

• temp
  Enables the use of a temporary working directory during execution (True or False).

• tmp_dir
  Path to the temporary directory used for intermediate files and data.

Parameters defining the electronic band structure of the material.

• band_gap
  Energy gap between the valence and conduction bands.

• m_e
  Effective mass of electrons in the conduction band.

• m_h
  Effective mass of holes in the valence band.

Time-evolution and initial-state parameters.

• num_steps
  Number of discrete time steps used in the dynamical simulation.

• evolution_time
  Total duration of the time evolution.

• dephasing_energy
  Energy scale associated with dephasing processes.

• init_cb
  Initial population of the conduction band.

• init_vb
  Initial population of the valence band.

External driving pulse parameters.

• t0
  Center time of the pulse.

• sigma
  Temporal width of the pulse.

• amp
  Pulse amplitude.

Execution and performance-related settings.

• shots
  Number of measurement shots used for sampling-based simulations.

• blocking_enable
  Enables qubit blocking to reduce circuit width.

• blocking_qubits
  Number of qubits per block when blocking is enabled.

• cuStateVec_enable
  Enables NVIDIA cuStateVec acceleration when available.

• optimization_level
  Circuit optimization level applied by the compiler/backend.

Quantum backend configuration.

• method
  Simulation method used by the backend.
  Available options:

  • statevector
  • matrix_product_state

• device
  Execution device.
  Available options:

  • CPU
  • GPU

• batched_shots_gpu
  Enables batched execution of shots on GPU backends.

Noise modeling configuration.

• noise
  Enables or disables noise in the simulation.

• model
  Noise model selection.
  Available options:

  • bitflip
  • frombackend (uses the backend’s native noise model)

• All parameters are read at runtime and can be modified to explore different physical regimes or backend configurations.
• For reproducibility, keep a copy of the exact config.ini file used for each simulation run.
• Backend- and noise-related options may depend on the availability of specific simulators or hardware.

This project is licensed under the MIT License. See the LICENSE file for details.