Qiskit — Open-Source Quantum Computing SDK by IBM

Introduction

Qiskit is IBM's open-source SDK for quantum computing. It lets developers create quantum circuits, simulate them locally, and execute them on real IBM quantum hardware through the cloud. The framework covers the full stack from circuit construction to transpilation and results analysis.

What Qiskit Does

• Constructs quantum circuits with gates, measurements, and classical control flow
• Transpiles circuits to match the topology and gate set of target hardware
• Runs circuits on local simulators or IBM Quantum cloud processors
• Provides a library of pre-built quantum algorithms (VQE, QAOA, Grover)
• Integrates with classical optimization and machine learning workflows

Architecture Overview

Qiskit is built around a core circuit model written in Rust for performance, exposed through a Python API. The transpiler converts abstract circuits into hardware-compatible instructions using a series of optimization passes. A primitives-based execution model (Sampler and Estimator) abstracts the difference between simulators and real devices. Results return as quasi-probability distributions or expectation values.

Self-Hosting & Configuration

• Install with pip install qiskit (Python 3.9+)
• For IBM hardware access, create a free IBM Quantum account and save your API token
• Use qiskit-aer for high-performance local simulation with GPU support
• Configure transpiler optimization levels (0-3) to trade compilation time for circuit quality
• Run Jupyter notebooks with qiskit[visualization] for interactive circuit diagrams

Key Features

• Rust-accelerated transpiler with multiple optimization passes
• Primitives API (Sampler/Estimator) for hardware-agnostic execution
• Built-in noise models for realistic simulation
• Qiskit Runtime for executing parameterized circuits efficiently on cloud hardware
• Extensive visualization tools for circuits, Bloch spheres, and histograms

Comparison with Similar Tools

• Cirq — Google's quantum SDK; lower-level with focus on NISQ devices, less integrated runtime
• PennyLane — Quantum ML focus with auto-differentiation; Qiskit is broader in scope
• Amazon Braket SDK — Access to multiple QPU vendors; Qiskit is deeper on IBM hardware
• Rigetti pyQuil — Tight Rigetti hardware integration; Qiskit has a larger community and algorithm library

FAQ

Q: Do I need quantum hardware to use Qiskit? A: No. Qiskit includes simulators for local development. Real hardware access is available free through IBM Quantum.

Q: What programming language does Qiskit use? A: Python for the user-facing API, with performance-critical internals written in Rust.

Q: Can Qiskit run on GPUs? A: Yes. The qiskit-aer simulator supports GPU acceleration via CUDA for large circuit simulations.

Q: Is Qiskit suitable for production quantum applications? A: Qiskit is used in research and early industry applications. Current quantum hardware is still noisy and limited in qubit count.

Sources

• https://github.com/Qiskit/qiskit
• https://docs.quantum.ibm.com