Simphony

Simphony is an open-source Python package from Qutility for simulating spin dynamics in central spin systems, with a focus on nitrogen-vacancy (NV) centers coupled to nuclear spins. It provides a modular and efficient framework for building quantum registers, designing pulse sequences, and analyzing system dynamics for quantum information and sensing applications.

Key features

• Build central spin registers by adding spins, interactions, and external fields.
• Simulate time evolution under pulse sequences to obtain the full unitary operator.
• Compute and visualize expectation values for chosen operators and initial states.
• Calculate process matrices in multiple bases and frames.
• Evaluate average gate fidelity against ideal operations.
• Include local quasi-static noise models to study error effects.
• Support simulations with multiple NV centers.
• Includes a predefined NV-center model with hyperfine interactions based on the Ivády Group's hyperfine dataset.

Technical specifications

• Uses internal time-evolution solvers based on NumPy and JAX.
• Supports both CPU and GPU backends.
• Accelerated by Just-in-Time (JIT) compilation via XLA.
• Enables automatic differentiation for gradient-based pulse optimization.

Future plans

• Add a predefined model for Silicon Carbide (SiC) with hyperfine interactions.
• Integrate photophysics dynamics, including initialization and readout.
• Include Lindblad-type noise models.

Requirements

• Python >=3.9
• CUDA 12 for GPU support

How to check your Python and CUDA versions:

python3 --version
nvidia-smi | grep CUDA

Installation

Simphony runs on CPU by default, but can achieve significant speedups with GPU acceleration. To enable GPU support, you need an NVIDIA GPU and a CUDA 12 environment. For more details, see the JAX GPU installation guide.

Install directly from GitHub

You can install the package directly from GitHub:

python3 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install "git+https://github.com/faulhornlabs/simphony.git"

With GPU (CUDA 12) support:

pip install "simphony[cuda12] @ git+https://github.com/faulhornlabs/simphony.git"

Installing with pip install git+... gives you the package only. The tutorial notebooks and local documentation sources are not kept in your working directory.

Clone and install locally

If you want the full repository, including jupyternbs/, clone and install it manually:

git clone https://github.com/faulhornlabs/simphony.git
cd simphony
python3 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install .

Editable install for development:

pip install -e .

Optional extras:

• cuda12: GPU-enabled JAX support for NVIDIA CUDA 12 environments.

  pip install ".[cuda12]"

• notebook-tools: Dependencies for running the tutorial notebooks locally.

  pip install ".[notebook-tools]"

• docs: Dependencies for building the documentation locally.

  pip install ".[docs]"

Combine them as needed:

pip install ".[notebook-tools,cuda12]"

Usage

Usage of Jupyter Notebook or JupyterLab is highly recommended to explore the functionality of the package. Tutorial notebooks can be found within the jupyternbs directory.

1. Start the JupyterLab by running: jupyter-lab
2. Select a notebook from the left panel within the pop-up browser window. For a first example, open jupyternbs/1_tutorial_basic.ipynb.

Documentation

Simphony documentation is available here