Explore the science, math, and engineering behind quantum computers — from qubits and gates to quantum machine learning.
01 — Computing Basics
Before diving into advanced quantum mechanics, understand the foundational principles that separate quantum from classical computation.
Unlike a classical bit (0 or 1), a qubit can exist in a superposition of both states simultaneously. Written as |ψ⟩ = α|0⟩ + β|1⟩, where α and β are complex probability amplitudes satisfying |α|² + |β|² = 1.
A qubit in superposition exists in multiple states at once until measured. This allows quantum computers to explore many solutions simultaneously — exponentially more efficient for certain problem classes.
Two or more qubits can become entangled, meaning the state of one instantly determines the state of another — regardless of the distance separating them. A crucial resource for quantum algorithms.
Quantum interference is used to amplify correct answers and cancel out wrong ones. Quantum algorithms are carefully designed so that paths leading to wrong answers destructively interfere.
Observing a qubit collapses its superposition to a definite 0 or 1. The probability of each outcome is determined by the squared magnitude of its amplitude: P(0) = |α|², P(1) = |β|².
Analogous to classical logic gates, quantum gates are unitary matrices that transform qubits. Common gates include Hadamard (H), Pauli-X/Y/Z, CNOT, and the Toffoli gate.
02 — Mathematics
The mathematical machinery underpinning quantum computation — from state vectors to error bounds.
Any qubit state is a unit vector in a 2D complex Hilbert space ℂ². The coefficients α and β are complex amplitudes; their squared magnitudes give measurement probabilities.
Every pure qubit state maps to a point on the unit sphere S² using polar angle θ ∈ [0, π] and azimuthal angle φ ∈ [0, 2π). Mixed states reside inside the sphere.
The Hadamard gate transforms |0⟩ → (|0⟩+|1⟩)/√2 and |1⟩ → (|0⟩−|1⟩)/√2, placing qubits in uniform superposition. It is its own inverse: H² = I.
Flips the target qubit t if and only if the control qubit c is |1⟩. Together with single-qubit gates, CNOT forms a universal gate set for quantum computation.
The simplest maximally entangled two-qubit state. Measuring one qubit instantly determines the other's outcome. The four Bell states form an orthonormal basis for ℂ² ⊗ ℂ².
The quantum analogue of the discrete Fourier transform, performed in O(n²) gates vs O(n·2ⁿ) classically. Core subroutine of Shor's algorithm and phase estimation.
The Schrödinger equation governs how quantum states evolve over time under a Hamiltonian Ĥ. Solution: |ψ(t)⟩ = e^(−iĤt/ℏ)|ψ(0)⟩, a unitary transformation.
Fidelity measures how similar two quantum states ρ and σ are. F = 1 means identical states; F = 0 means orthogonal states. Critical for benchmarking quantum hardware.
Measures the quantum entanglement and mixedness of a state. Zero for pure states, maximally log₂(d) for the d-dimensional maximally mixed state. Quantum analogue of Shannon entropy.
03 — Glossary
A reference guide to the essential vocabulary of quantum computing.
The quantum bit — the fundamental unit of quantum information. Unlike classical bits, a qubit can exist in a continuous superposition of |0⟩ and |1⟩ states until measured, enabling quantum parallelism.
A reversible quantum operation represented by a unitary matrix U (satisfying U†U = I). Gates manipulate qubits analogously to classical logic gates. Single-qubit gates rotate vectors on the Bloch sphere.
The process by which a quantum system loses its quantum properties through unwanted interactions with its environment. Decoherence is the primary challenge in building practical quantum computers and limits computation time.
The demonstration that a quantum device can solve a specific problem exponentially faster than any classical computer. First claimed by Google in 2019 with Sycamore (53 qubits, 200 seconds vs. ~10,000 years classically).
Techniques to protect quantum information from errors due to decoherence and imperfect gates. Unlike classical error correction, quantum states cannot be directly copied (no-cloning theorem), requiring indirect syndrome measurement strategies.
A metaheuristic approach to finding global optima in complex optimization problems using quantum fluctuations (tunneling) rather than thermal fluctuations. Employed by D-Wave systems for combinatorial optimization tasks.
(VQE) A hybrid quantum-classical algorithm that uses a parameterized quantum circuit to prepare trial states, then classically optimizes parameters to minimize the expectation value of a Hamiltonian. Central to NISQ-era quantum chemistry.
Noisy Intermediate-Scale Quantum devices — today's quantum hardware with 50–1000+ qubits but without full fault-tolerance. Research focuses on finding practical quantum advantage within these noisy constraints.
A protocol to transmit an arbitrary quantum state between two parties using a pre-shared entangled pair and two classical bits of communication. It does not transmit information faster than light; classical communication is still required.
04 — Applications
Where quantum computing creates transformative impact — today and in the near future.
Simulating molecular interactions at quantum scale allows exact modeling of drug-protein binding. Quantum computers can simulate the nitrogen-fixing enzyme FeMoco using ~200 logical qubits — impossible for classical supercomputers.
Life SciencesShor's algorithm threatens RSA/ECC encryption. This drives development of post-quantum cryptography (lattice-based, hash-based schemes). NIST finalized the first PQC standards in 2024.
CybersecurityQuantum algorithms like QAOA can solve quadratic unconstrained binary optimization (QUBO) problems, enabling faster portfolio rebalancing, risk modeling, options pricing, and fraud detection at scale.
FinanceDesigning better solar cells, room-temperature superconductors, and CO₂ catalysts requires quantum simulation. Classical density functional theory fails at correlated electron systems where quantum methods excel.
EnergyQuantum optimization solves NP-hard problems like the Travelling Salesman Problem exponentially faster for large instances. Volkswagen and DHL have piloted quantum routing on real traffic data.
LogisticsQuantum kernels can map classical data into exponentially large feature spaces, potentially offering advantages in classification, clustering, and generative modelling — especially for structured quantum data.
Quantum ML05 — Quantum Machine Learning
Quantum Machine Learning (QML) fuses quantum computing with classical ML — unlocking new representations, training dynamics, and model expressibility.
↑ Parameterized Quantum Circuit (PQC) — the quantum analogue of a neural network layer
Parameterized quantum circuits used as trainable models. Gradients computed via the parameter-shift rule: ∂⟨O⟩/∂θ = ½[⟨O⟩(θ+π/2) − ⟨O⟩(θ−π/2)]. Trainable with classical optimizers (Adam, BFGS).
Map classical data x into quantum feature states |φ(x)⟩ via an encoding circuit. The quantum kernel k(x,x') = |⟨φ(x')|φ(x)⟩|² can capture correlations in exponentially large Hilbert spaces inaccessible to classical kernels.
Quantum Boltzmann Machines and Quantum GANs use quantum states to represent complex probability distributions. Born machines directly encode distributions as the squared amplitudes of quantum states, enabling efficient sampling.
A major challenge: gradients of deep random quantum circuits vanish exponentially with qubit count, making training infeasible. Solutions include local cost functions, problem-specific ansätze, and layer-wise training strategies.
NISQ algorithms like VQE and QAOA use a classical optimizer to update quantum circuit parameters based on measurement outcomes. The quantum device evaluates the cost function; the classical computer handles gradient descent.
06 — Algorithms
The algorithms that define quantum advantage — their speedups, and where they apply.
| Algorithm | Classical Complexity | Quantum Complexity | Speedup | Primary Application | Type |
|---|---|---|---|---|---|
| Shor's Algorithm | O(e^(n^(1/3))) | O(n³) | Exponential | Integer Factorization, RSA Breaking | Cryptography |
| Grover's Algorithm | O(N) | O(√N) | Quadratic | Unstructured Database Search | Search |
| HHL Algorithm | O(N·κ²/ε) | O(log(N)·κ²/ε) | Exponential* | Linear Systems Ax = b | Linear Algebra |
| VQE | O(2^n / ε) | Heuristic NISQ | Practical | Ground State Energy, Chemistry | Simulation |
| QAOA | Exponential | Heuristic NISQ | Heuristic | Combinatorial Optimization (MaxCut) | Optimization |
| Quantum Phase Estimation | O(1/ε²) | O(1/ε) | Quadratic | Eigenvalue Estimation, QFT Core | Subroutine |
| Bernstein-Vazirani | O(n) queries | O(1) query | Exponential | Hidden String Problem, Oracles | Query |
| Quantum Walk | O(N) | O(√N) | Quadratic | Graph Traversal, Element Distinctness | Graph |
07 — History
From theoretical concept to a new computing paradigm — the milestones that shaped the field.
Richard Feynman proposes that a computer built on quantum mechanical principles could simulate quantum systems far more efficiently than any classical machine.
David Deutsch formalizes the quantum Turing machine and demonstrates the first quantum algorithm, introducing the concept of quantum parallelism.
Peter Shor's polynomial-time quantum algorithm for integer factorization triggers worldwide interest in quantum computing — threatening RSA encryption.
Lov Grover presents a quantum algorithm achieving quadratic speedup for unstructured search, proving quantum advantage beyond simulation tasks.
IBM demonstrates Shor's algorithm on a 7-qubit NMR quantum computer, factoring 15 = 3 × 5 — a landmark experimental validation.
D-Wave releases the first commercial quantum annealer with 128 qubits, sparking debate about quantum advantage in optimization problems.
IBM opens cloud access to a 5-qubit quantum processor — democratizing quantum computing and launching the quantum cloud era.
Google's Sycamore processor (53 qubits) performs a specific sampling task in 200 seconds that would take classical supercomputers an estimated 10,000 years.
IBM unveils Condor with over 1,000 superconducting qubits alongside the Heron processor, advancing scalable quantum architecture toward fault-tolerant computing.
Microsoft, Google (Willow), and IBM make significant strides in logical qubit demonstration and error correction, targeting the first fault-tolerant quantum computers by the late 2020s.