QONTOS Simulators

Simulation, architecture estimation, and tensor-network modeling for the QONTOS platform.

Public validation and planning tools for the software stack today and the modular hardware roadmap ahead.

Overview · Installation · Quick Start · Docs Hub · Simulators · Architecture Estimator · Tensor Engine · Related Packages

---

Overview

QONTOS Simulators is a small, self-contained SDK for building and running quantum algorithms. The whole package depends only on NumPy: a friendly circuit builder, an exact statevector simulator, and a matrix-product-state (tensor-network) backend for larger, low-entanglement circuits. It also ships a modular-architecture ESTIMATOR (a planning sandbox, not a measured-data-calibrated digital twin) for system-level planning; see MODEL_CARD.md for its valid domain.

Start with docs/index.md for the lightweight docs hub.

It provides:

1. qontos_sim — the developer SDK: a Circuit builder and one simulate call, with an exact statevector backend and a tensor-network (MPS) backend.
2. qontos_twin — a modular-architecture ESTIMATOR (planning sandbox) for architecture and throughput studies. Not a measured-data-calibrated digital twin; see MODEL_CARD.md.
3. qontos_tensor — a pure NumPy tensor-network engine (MPS, MPO, DMRG) that powers the MPS backend.

Installation

Requires Python 3.10+ and NumPy. Nothing else.

Pre-release (current)

Not yet on PyPI. Install from source or a pinned release tag:

pip install "qontos-sim @ git+https://github.com/qontos/qontos-sim.git@v0.1.0"

Once published to PyPI this becomes pip install qontos-sim.

Quick Start

Build a Bell pair and run it in three lines:

from qontos_sim import Circuit, simulate

c = Circuit(2)
c.h(0).cx(0, 1).measure_all()
result = simulate(c, shots=1000)
print(result.counts)          # {'00': ~500, '11': ~500}

A 3-qubit GHZ state on the tensor-network backend:

from qontos_sim import Circuit, simulate

ghz = Circuit(3).h(0).cx(0, 1).cx(1, 2).measure_all()
print(simulate(ghz, shots=1000, method="mps").counts)   # {'000': ~500, '111': ~500}

Bitstring convention: position i is the measured value of qubit i (qubit 0 leftmost).

Architecture Estimator

from qontos_twin import ModularSimulator, SystemConfig

config = SystemConfig(
    num_modules=4,
    transduction_efficiency=0.15,
)
sim = ModularSimulator(config)
workload = sim.simulate_workload(circuit_depth=250)
print(f"Estimated fidelity: {workload.estimated_fidelity:.4f}")
print(f"Bell pairs required: {workload.bell_pairs_needed}")

Tensor Network Simulation

from qontos_tensor import GateInstruction, TNSimulator

# Simulate bounded-entanglement circuits with an MPS backend
sim = TNSimulator(n_qubits=2, chi_max=256)
result = sim.run(
    [
        GateInstruction(name="H", qubits=[0]),
        GateInstruction(name="CNOT", qubits=[0, 1]),
    ],
    n_shots=1024,
)
print(result.measurements[:5])

Simulators

How to call	Backend	Qubits	Use case
simulate(c, method="statevector")	Exact statevector (NumPy)	Up to ~25	The default; exact, any circuit
simulate(c, method="mps")	Tensor network (MPS)	Larger, bounded entanglement	Big low-entanglement circuits
ModularSimulator (qontos_twin)	Architecture estimator (planning sandbox)	Unlimited (modeled)	Architecture and throughput studies

Architecture Estimator

The architecture estimator analyses workloads on modular architecture candidates using scenario bands and analytic proxies. It is a planning signal, not a measured-fidelity prediction (see MODEL_CARD.md). For a given system configuration it estimates:

• Total gate count (intra-module and inter-module)
• Circuit fidelity (based on gate fidelity, transduction, and decoherence)
• Runtime in microseconds
• Bell pairs required for inter-module operations
• Effective circuit depth increase from serialization

Transduction Scenario Bands

Efficiency	Scenario	Interpretation
>= 20%	Stretch	Full modular planning
>= 10%	Aggressive	Meaningful multi-module operation
1-10%	Base	Staged modular validation
< 1%	Research	Device and link R&D

Tensor Network Engine

Pure NumPy implementation — zero external tensor network dependencies.

• MPS (Matrix Product State) — Bond dimension up to 4096
• MPO (Matrix Product Operator) — Heisenberg, Ising, Hubbard, molecular Hamiltonians
• DMRG — Variational ground-state search for 100+ site systems
• Circuit simulation — Full circuit evolution via MPS

Examples

Runnable scripts live in examples/:

• quickstart.py — Bell and GHZ states end to end.
• chsh_bell_inequality.py — reproduces the CHSH violation (S approaches 2.83, beating the classical bound of 2).
• variational_sweep.py — a one-qubit energy sweep, the kernel of a variational algorithm.

Related repositories

Repository	Description
qontosq	The QONTOS quantum SDK for the modular architecture: cross-module transpilation to Bell pairs plus feed-forward, link noise, and a QEC-aware logical layer
qontos-examples	Verified, CI-executed examples against the pinned public releases

More of the ecosystem (benchmarks, research) will be linked as it is published.

License

Apache License 2.0

---

Built by Zhyra Quantum Research Institute (ZQRI) — Abu Dhabi, UAE