tsim: fast sampling of Clifford+T circuits

tsim is a quantum circuit sampler designed for efficient sampling of Clifford+T circuits with Pauli noise.

tsim follows the stim API and works with stim circuits. It supports all stim gates and noise channels, and, importantly, T-gates.

Quick Start

c = tsim.Circuit(
    """
    RX 0
    T 0
    DEPOLARIZE1(0.1) 0
    H 0
    M 0
    """
)
sampler = c.compile_sampler()
samples = sampler.sample(shots=100)

For circuits with detector and observable annotations, you can compile a detector sampler:

c = tsim.Circuit(
    """
    RX 0
    R 1
    T_DAG 0
    PAULI_CHANNEL_1(0.1, 0.1, 0.2) 0 1
    H 0
    CNOT 0 1
    M 0 1
    DETECTOR rec[-1] rec[-2]
    """
)
detector_sampler = c.compile_detector_sampler()
samples = detector_sampler.sample(shots=100)

Installation

uv add bloqade-tsim

For GPU acceleration, use

uv add "bloqade-tsim[cuda13]"

See Installation for more options.

How It Works

tsim uses stabilizer rank decomposition based on the ZX calculus and is built on top of pyzx. Circuits are converted into ZX diagrams where noise channels are injected as parametrized Pauli vertices. For efficient sampling on CPU and GPU, the diagram is compiled into contiguous jax arrays, following the approach described in arXiv:2403.06777.