Tensorcircuit Ng
Tensorcircuit Ng: next-gen tensor-network-based quantum software framework with JAX/TensorFlow/PyTorch backends.
Last verified:
What is Tensorcircuit Ng?
Tensorcircuit Ng (TensorCircuit-NG) is an open-source high-performance quantum computing software framework in Python designed for the NISQ (Noisy Intermediate-Scale Quantum) era. It provides a comprehensive set of tools for quantum circuit simulation with support for multiple backends including Numpy, TensorFlow, JAX, and PyTorch. The framework is empowered by advanced tensor network simulator engines and is ready for quantum hardware access with CPU/GPU/QPU local and cloud hybrid solutions.
Key features include unified quantum programming across Jax/TensorFlow/PyTorch/Numpy/Cupy backends, support for CPU/GPU/TPU hardware, compatibility with QPUs from different vendors, and the ability to build and simulate tensor networks, neural networks, and quantum circuits together. It supports automatic differentiation, just-in-time compilation, vectorized parallelism, and GPU acceleration, making it compatible with machine learning engineering paradigms. The framework also supports ideal, noisy, approximate, and analog simulation modes, and can convert to/from IR formats like Qiskit, OpenQASM, and JSON.
Tensorcircuit Ng is designed for quantum computing researchers, quantum algorithms developers, academic researchers working on quantum many-body problems, and industry professionals deploying quantum computing solutions. It helps users efficiently solve challenging quantum computing problems from academic research prototypes to industry application deployment. The framework is maintained by Tencent Quantum Lab and is built for humans with a focus on speed, flexibility, and elegance.
Tensorcircuit Ng pricing
Pricing model: Freemium
TensorCircuit-NG is completely free and open-source. It is distributed via PyPI as the tensorcircuit-ng package and can be installed using pip install. There are no paid tiers or commercial licensing requirements. Optional extra dependencies are available for specific backends (tensorflow, jax, torch, qiskit, cloud) but remain free open-source packages.
Tensorcircuit Ng pros
- Open-source and free to use
- Supports multiple ML backends (TensorFlow, JAX, PyTorch, Numpy)
- Advanced tensor network simulator engines
- GPU and TPU acceleration support
- Automatic differentiation built-in
- Just-in-time compilation for performance
- Works with real quantum hardware (QPU access)
- Local and cloud hybrid solutions
- Compatible with Qiskit and OpenQASM
- Vectorized parallelism support
- Supports noisy and ideal simulation modes
- Unified interface across different backends
- Jupyter notebook tutorials available
- Extensible architecture for research
- Industry-ready for deployment
Tensorcircuit Ng cons
- Steep learning curve for quantum computing beginners
- Requires understanding of tensor network concepts
- Optional backend dependencies add complexity
- Documentation primarily in English
- Relatively new compared to established frameworks
- May require GPU for large-scale simulations
- Limited pre-built quantum algorithms library
- Active development may have breaking changes
Frequently asked questions about Tensorcircuit Ng
What is TensorCircuit-NG?
TensorCircuit-NG is an open-source high-performance quantum software framework in Python designed for the NISQ era. It is the next generation of TensorCircuit, built for speed, flexibility and elegance, empowered by advanced tensor network simulator engines, and implemented with industry-standard machine learning frameworks including TensorFlow, JAX, and PyTorch.
What backends does TensorCircuit-NG support?
TensorCircuit-NG supports multiple backends including Numpy, TensorFlow, JAX, PyTorch, and Cupy. It also supports CPU, GPU, and TPU hardware acceleration, and can work with QPUs from different vendors through its unified quantum programming interface.
How do I install TensorCircuit-NG?
TensorCircuit-NG can be installed via pip using 'pip install tensorcircuit-ng'. It is also available via Docker with 'docker pull'. Core dependencies include numpy, scipy, tensornetwork-ng, and networkx. Optional backend dependencies like tensorflow, jax, jaxlib, torch, and qiskit can be installed separately.
What simulation modes are available?
TensorCircuit-NG supports ideal simulation, noisy simulation, approximate simulation, and analog simulation modes. This flexibility allows users to choose the appropriate simulation type based on their research needs and available computational resources.
Can I access real quantum hardware with TensorCircuit-NG?
Yes, TensorCircuit-NG is ready for quantum hardware access with CPU/GPU/QPU hybrid solutions supporting both local and cloud deployments. It can interface with QPUs from different vendors through its unified quantum programming infrastructure.
Does TensorCircuit-NG support automatic differentiation?
Yes, TensorCircuit-NG is fully compatible with machine learning engineering paradigms including automatic differentiation via tc.backend.grad(). This allows users to compute gradients of quantum circuits for optimization and machine learning applications.
What is the difference between TensorCircuit and TensorCircuit-NG?
TensorCircuit-NG is the next-generation version that supersedes the original TensorCircuit (v0.x). It is built with improved performance, better flexibility, and enhanced elegance. The NG version has updated core authors and maintains backward compatibility while offering improved tensor network simulator engines.
Are there tutorials available for TensorCircuit-NG?
Yes, TensorCircuit-NG provides Jupyter Notebook tutorials at the documentation site. There are also educational resources including Quantum Computing Lectures with TC-NG available on GitHub. Example scripts in the /examples/ directory serve as integration tests and learning resources.
Can TensorCircuit-NG convert to Qiskit or OpenQASM?
Yes, TensorCircuit-NG supports conversion to and from IR formats including Qiskit, OpenQASM, and JSON. This interoperability allows users to integrate TensorCircuit-NG circuits with other quantum computing frameworks and quantum hardware platforms.
Who maintains TensorCircuit-NG and how do I get support?
TensorCircuit-NG is created and maintained by Tencent Quantum Lab. The current core authors can be found in the documentation. For questions or collaboration ideas, users can use the GitHub Issues tracker, GitHub Discussions forum, or send email to [email protected].