NISO - NISQ Integrated System Optimizer

High-performance Rust implementation of TQQC (Temporal Noise Quantum Computing) optimization algorithms for NISQ devices.

Current Version: 1.2 (Phase 5: IBM Hardware Integration Completed)

Features

• TQQC v2.2.0: Dynamic inner loop, adaptive z-test, early stopping (~12% avg improvement).
• IBM Quantum Integration: Direct hardware execution via niso_qiskit (Native Rust implementation).
• Multi-target Hardware: Superconducting (IBM), Trapped Ion, Neutral Atom support.
• Python Bindings: Seamless Python API via niso package (PyO3).
• Comprehensive Benchmarking: Noise/qubit/points scaling analysis tools.

Architecture

NISO follows a modular layered architecture (L0-L10):

NISO_System_v1.2
├── niso_core        # L0-L1: Foundation types, Circuit builder
├── niso_noise       # L2: Noise modeling (T1, T2, Errors)
├── niso_calibration # L3: Hardware calibration management
├── niso_schedule    # L4: Circuit scheduling & decoherence estimation
├── niso_tqqc        # L5: TQQC optimization engine
├── niso_backend     # L6: Execution interface (Simulator/Hardware)
├── niso_engine      # L7: High-level integration pipeline
├── niso_bench       # L8: Benchmark suites
├── niso_python      # L9: Python bindings
└── niso_qiskit      # L10: IBM Quantum Job/API management

Quick Start

1. Robust Optimization (Simulation)

use niso_engine::prelude::*;

// Create optimizer with 7 qubits and 2% noise
let config = NisoConfig::default_7q()
    .with_noise(0.02)
    .with_seed(42);

let mut optimizer = NisoOptimizer::new(config)?;
let result = optimizer.optimize()?;

println!("Improvement: {:.2}%", result.improvement_percent());

2. IBM Quantum Real Hardware

export IBM_QUANTUM_TOKEN="your-api-token"

use niso_qiskit::prelude::*;
use niso_core::CircuitBuilder;

// Connect to IBM Brisbane
let backend = IbmBackend::from_env("ibm_brisbane")?;

// Execute circuit (Automatic transpilation included)
let circuit = CircuitBuilder::new(127)
    .h(0)
    .cx(0, 1)
    .measure_all()
    .build();

let result = backend.execute_sync(&circuit, 4096)?;
println!("Parity: {:.4}", result.parity_expectation());

3. Python Usage

import niso

# Quick optimization
result = niso.quick_optimize(qubits=7, noise=0.02, seed=42)
print(f"Improvement: {result.improvement_percent:.2f}%")

# Generate benchmark report
suite = niso.BenchSuite(seed=42)
suite.noise_scaling(7, [0.01, 0.02, 0.03], 10)

Installation

Rust

git clone https://github.com/sadpig70/NISO.git
cd niso
cargo build --release
cargo test

Python

pip install maturin
cd crates/niso_python
maturin develop --release

Performance & Results

TQQC v2.2.0 on 7-qubit H-chain (Simulation):

Metric	Result	Note
Avg Improvement	12.13%	p=0.02 (Significant)
Max Improvement	19.82%	p=0.015
Computation Saved	~51%	Early stopping & dynamic loops

Test Coverage (v1.1):

• Total Tests: 350
• Passing: 350 (3 Ignored)
• Includes full integration tests for IBM Qiskit API.

Project Structure

niso/
├── Cargo.toml          # Workspace config
├── README.md           # This file
├── docs/               # Documentation
│   ├── NISO_Unified_Specification_v1.2.md # Unified Spec
│   └── ...

└── crates/             # Source Code
    ├── niso_core/
    ├── niso_noise/
    ├── niso_calibration/
    ├── niso_schedule/
    ├── niso_tqqc/
    ├── niso_backend/
    ├── niso_engine/
    ├── niso_bench/
    ├── niso_python/
    └── niso_qiskit/    # [New] IBM Support

License

MIT

Author

Jung Wook Yang (sadpig70@gmail.com) QC Technical Partner