Foundations of Quantum Computing Applications

Introduction and context

Quantum computing is a method of information processing that uses physical systems governed by quantum mechanics. Its importance does not arise simply because quantum computers are faster versions of classical computers. Rather, quantum processors represent and transform information in ways that have no direct classical analogue. These capabilities may provide advantages for selected problems, including simulation of quantum systems, some search and optimisation tasks, cryptanalysis, and certain sampling and machine-learning workloads.

A useful foundation must distinguish four related but different fields:

• Quantum computing uses quantum states to perform computations.
• Quantum communication uses quantum states to transmit information or establish security properties.
• Quantum sensing uses quantum effects to measure physical quantities with exceptional sensitivity.
• Classical high-performance computing (HPC) uses conventional bits, processors, accelerators and networks, often at very large scale.

Quantum technology is therefore an ecosystem rather than a single machine category. A quantum computer may work alongside classical processors, while communication and sensing systems may use quantum devices without executing general-purpose quantum algorithms.

Learning objectives

After studying this chapter, a learner should be able to:

1. Explain the physical and mathematical meaning of a qubit.
2. Describe superposition, entanglement, interference and measurement.
3. Interpret common quantum gates and simple quantum circuits.
4. Explain why quantum algorithms are probabilistic and how results are validated.
5. Relate quantum resources to computational complexity and speedup claims.
6. Distinguish quantum computing from classical HPC, quantum communication and quantum sensing.
7. Identify practical applications, limitations, governance issues and professional best practices.

From classical bits to qubits

A classical bit has one of two logical values: 0 or 1. A physical implementation may use voltage ranges, magnetic orientations or transistor states. A classical register of n bits contains one definite n-bit string at a time, although a computer can process many strings over a sequence of operations.

A qubit is a two-level quantum system. Its computational-basis states are written as |0⟩ and |1⟩. Before measurement, a pure qubit may be represented by

Here α and β are generally complex probability amplitudes. The normalization condition is

The quantities |α|² and |β|² are real probabilities. If the qubit is measured in the computational basis, the probability of obtaining 0 is |α|² and the probability of obtaining 1 is |β|². The amplitudes themselves are not probabilities; their phases can affect interference.

A convenient parameterisation of a pure qubit is

where θ and φ are real angles. This representation assumes a pure state and omits a physically irrelevant global phase. The Bloch sphere is a visual model for this state space: θ determines latitude and φ determines azimuthal phase.

A superposition is not the same as a classical random choice. A classical random bit is in either 0 or 1, with uncertainty about which value occurred. A coherent quantum superposition contains amplitudes with relative phase information. That phase may later produce constructive or destructive interference.

Multi-qubit states

For two qubits, the computational basis consists of |00⟩, |01⟩, |10⟩ and |11⟩. A general pure state is

with

For n qubits, the state vector has 2ⁿ amplitudes. This exponential state-space dimension is important, but it does not mean that a quantum computer can directly read 2ⁿ classical answers. Measurement yields only one outcome per circuit execution. Quantum algorithms must arrange amplitudes so that useful outcomes become more likely.

Superposition, entanglement and interference

Superposition

The Hadamard gate illustrates superposition. Its matrix is

Using the computational basis, it transforms

and

Thus, measuring H|0⟩ gives 0 or 1 with equal probability. However, applying H a second time returns |0⟩, because the relative signs interfere. This simple example shows why “trying both values at once” is an incomplete description of quantum computation: the useful operation is controlled manipulation of amplitudes and phases.

Entanglement

A multi-qubit state is entangled when it cannot be expressed as a tensor product of individual qubit states. The Bell state

is entangled. Measuring either qubit gives 0 or 1 with equal probability, but the two results are always equal when measured in the same basis. Entanglement creates correlations that cannot be reproduced by assigning independent local quantum states to the two qubits.

Entanglement does not permit faster-than-light communication. The individual result is random, and comparison of results requires an ordinary classical communication channel. In computing, entanglement is a resource used by algorithms, error-correction schemes and communication protocols.

Interference

If two computational paths lead to the same outcome, their complex amplitudes add. The probability is the squared magnitude of the total amplitude:

For amplitudes A₁ and A₂, this expands to

where Re denotes the real part and * denotes complex conjugation. The final term is the interference term. It may increase or decrease the probability. Quantum algorithms exploit this effect to amplify desirable outcomes and suppress undesirable ones.

Measurement and probabilistic output

Measurement maps a quantum state to a classical result. For a computational-basis measurement of |ψ⟩ = α|0⟩ + β|1⟩, the result is 0 with probability |α|² or 1 with probability |β|², and the post-measurement state is the corresponding basis state.

A circuit is normally executed repeatedly, producing a collection of outcomes called shots. If an event has probability p and is measured N independent times, its expected count is Np. The estimated frequency has standard deviation approximately

where f is the observed frequency. This binomial approximation assumes identical independent shots and a stable device. Increasing N reduces statistical uncertainty approximately as 1/√N, but it does not remove systematic device errors.

Measurement basis matters. A state that is definite in one basis may be a superposition in another. Gates can rotate the state before measurement, allowing different properties to be investigated.

Quantum gates and circuits

Ideal quantum gates are reversible and are represented by unitary matrices U satisfying

where U† is the conjugate transpose and I is the identity matrix. Unitary evolution preserves total probability. Measurement is not unitary and is usually treated as the final process in a circuit, although intermediate measurement and reset are also possible.

Common one-qubit gates include:

A controlled-NOT (CNOT) has a control and a target. It flips the target if the control is 1:

where a and b are bits and ⊕ denotes addition modulo 2. CNOT, together with suitable single-qubit gates, can create entanglement. For example, starting with |00⟩, applying H to the first qubit and CNOT with the first qubit as control gives

This produces the Bell state |Φ⁺⟩.

A circuit is a time-ordered description of operations. Wires represent qubits; gates act from left to right or, in some conventions, from top to bottom. Circuit depth is the number of sequential layers. Gate count measures the number of operations. Both affect execution time, noise exposure and compilation cost.

A small worked example: phase and interference

Consider the input |0⟩ and the sequence H–Z–H.

1. After the first H:

|0⟩ → (|0⟩ + |1⟩)/√2

1. After Z:

(|0⟩ + |1⟩)/√2 → (|0⟩ − |1⟩)/√2

1. After the final H, using H|0⟩ and H|1⟩:

(|0⟩ − |1⟩)/√2 → |1⟩

Therefore measurement returns 1 with probability 1 in the ideal, noiseless circuit. The Z gate did not change the computational-basis probabilities at the moment it was applied, but it changed the relative phase. The final H converted that phase difference into a measurable probability difference.

Computational complexity and quantum advantage

Computational complexity studies how resource requirements grow with input size. Common resources include time, memory, circuit depth, number of gates, communication and oracle queries. Big-O notation describes asymptotic upper-growth behaviour; for example, O(n²) indicates that a resource is bounded by a constant multiple of n² for sufficiently large n, ignoring lower-order terms and constants.

Important complexity classes include:

• P: decision problems solvable by a deterministic classical algorithm in polynomial time.
• BPP: problems solvable by a probabilistic classical algorithm with bounded error in polynomial time.
• BQP: decision problems solvable by a quantum computer in polynomial time with bounded error.
• NP: problems whose proposed solutions can be verified in polynomial time by a classical computer.

The relationship among these classes is not fully known. It is commonly understood that P ⊆ BPP ⊆ BQP, although whether any inclusion is strict is unresolved in general. NP is not known to be contained in BQP, and quantum computing is not known to efficiently solve all NP-complete problems.

A claimed quantum speedup must specify:

1. the problem and input model;
2. the classical and quantum algorithms being compared;
3. the relevant resource, such as runtime, queries or samples;
4. error tolerance and success probability;
5. hardware, data-loading and output costs.

Examples of important algorithmic ideas include Shor’s polynomial-time factoring algorithm under standard idealised assumptions, Grover’s quadratic query improvement for unstructured search, and quantum simulation methods for physical systems. A quadratic improvement is valuable but is not an exponential improvement. Furthermore, asymptotic advantage may not translate into practical advantage at small problem sizes or on noisy hardware.

Quantum computing compared with neighbouring fields

HPC and quantum computing are complementary. HPC offers mature software, large memory, fault tolerance and predictable performance. Quantum processors have limited connectivity, noise and measurement overhead, but may accelerate specific subroutines. A realistic architecture is hybrid: classical software prepares inputs and controls the quantum processor, while the quantum device executes a circuit and returns samples.

Quantum communication includes quantum key distribution, in which security is associated with physical principles and the detection of certain interception attempts. It does not mean that all communication becomes instantaneous or that a quantum computer is required at every endpoint.

Quantum sensing may use superposition, interference, squeezed states or entanglement to improve measurement. A quantum sensor measures a physical parameter; a quantum computer transforms encoded information to solve a computational problem. Some technologies overlap, but their engineering objectives and performance metrics differ.

Applications and professional workflow

Potential application areas include:

• Chemistry and materials: estimating molecular energies, reaction pathways and material properties.
• Optimisation: exploring structured search spaces in logistics, scheduling, routing and portfolio construction, subject to careful benchmarking.
• Cryptography: analysing the future impact of factoring and discrete-logarithm algorithms and migrating to post-quantum cryptography.
• Machine learning: investigating quantum kernels, variational circuits and sampling methods, while accounting for data-loading and training costs.
• Finance and risk: studying Monte Carlo-related procedures and portfolio constraints, with no automatic guarantee of business advantage.
• Scientific simulation: modelling quantum systems that are difficult to represent directly with classical resources.

A professional workflow begins with problem formulation rather than hardware selection. The analyst should establish whether the problem has exploitable structure, whether input data can be encoded efficiently, whether output precision is sufficient, and whether a classical baseline is strong. The workflow then proceeds through circuit design, compilation, resource estimation, simulator testing, hardware execution, statistical analysis and independent benchmarking.

Illustrative case: molecular-energy estimation

Background. A materials company wants to estimate the ground-state energy of a small molecule to screen candidate catalysts. This is an illustrative case, not a claim of current universal quantum advantage.

Problem. The company needs an energy estimate within a specified tolerance. Classical quantum-chemistry software already provides a baseline, but the team wishes to assess whether a hybrid quantum method could become useful as system size grows.

Analysis. The molecular Hamiltonian is mapped to a qubit Hamiltonian. A trial state is prepared by a parameterised circuit. Measurements estimate expectation values of the Hamiltonian terms. A classical optimiser updates circuit parameters.

For an observable H and state |ψ(θ)⟩, the objective is

The variational principle states that E(θ) is at least the ground-state energy when H is bounded below and |ψ(θ)⟩ is a normalised trial state. The expectation value is estimated from repeated measurements, so sampling noise and device noise affect optimisation.

Approach. The team first validates the model on a classical simulator, then compares a noise-aware simulation with hardware results. It reports circuit depth, number of shots, error mitigation, confidence intervals and the classical baseline. It does not describe a lower measured energy as proof of quantum advantage unless the comparison is fair and reproducible.

Outcome. The current device produces an informative demonstration but not a production-quality advantage. The exercise identifies requirements for future progress: improved qubit fidelity, shallower ansätze, better error mitigation and efficient measurement allocation.

Learning points. Application value depends on the entire workflow, not only on qubit count. Domain formulation, verification, uncertainty reporting and classical comparison are essential.

Noise, error correction and limitations

Real qubits interact with their environment. Important error mechanisms include relaxation, dephasing, gate errors, readout errors, crosstalk and leakage outside the computational subspace. Coherence times describe how long certain quantum information survives, but a longer coherence time alone does not guarantee useful computation.

Noise may be characterised by metrics such as gate fidelity, readout accuracy, error rates, circuit depth and quantum volume-like benchmarks. No single metric captures application performance. Error mitigation uses calibration, extrapolation, probabilistic techniques or symmetry checks to reduce observed error without fully correcting every error during computation. It generally increases sampling cost and does not replace fault-tolerant error correction.

Quantum error correction encodes logical qubits into multiple physical qubits. Fault tolerance aims to perform arbitrarily long computations when physical error rates are below relevant thresholds and engineering conditions are satisfied. The overhead can be substantial. Consequently, near-term noisy devices and future fault-tolerant computers should not be treated as interchangeable.

Other limitations include:

• difficulty connecting qubits and routing two-qubit gates;
• costly state preparation and classical data loading;
• limited precision in measured outputs;
• compiler and calibration drift;
• scarcity of skilled personnel;
• uncertain economic advantage for many proposed applications.

Governance, security and responsible practice

Quantum technologies create professional responsibilities. Organisations should avoid unsupported claims such as “quantum solves every optimisation problem” or “more qubits automatically means more power.” Benchmark definitions, baselines, error bars and resource assumptions should be published.

Cryptographic planning is especially important. Even before large fault-tolerant quantum computers exist, an adversary may collect encrypted data for later decryption. Organisations should inventory cryptographic dependencies, prioritise long-lived sensitive information and evaluate post-quantum cryptographic standards and migration plans. Quantum key distribution is not a universal replacement for authentication, secure endpoints or sound key management.

Responsible practice also includes protecting proprietary data sent to cloud quantum services, documenting whether experiments use simulators or hardware, reporting failed trials, and considering energy, supply-chain and access issues. Quantum software should be tested against classical alternatives and audited for reproducibility.

Contemporary development and best practices

Current development is characterised by hybrid quantum-classical algorithms, cloud access, improved control electronics, error mitigation, logical-qubit research and increasing attention to application-specific benchmarks. The field remains rapidly changing; roadmaps and marketing claims should therefore be separated from demonstrated capability.

Best practices for diploma-level projects are:

1. Define the application and success metric before selecting a circuit.
2. Build a transparent classical baseline.
3. Use small, interpretable circuits before increasing scale.
4. Track qubit mapping, depth, gate counts, shots and calibration conditions.
5. Separate algorithmic error, sampling error and hardware noise.
6. Repeat experiments and report uncertainty.
7. Verify results with simulators and alternative classical methods.
8. Treat quantum hardware as a component in a hybrid system.
9. Protect data and assess cryptographic implications.
10. Report limitations honestly.

Chapter Summary

Quantum computing processes information using qubits, whose amplitudes obey normalisation and whose measurement outcomes are probabilistic. Superposition provides coherent alternatives, entanglement creates non-classical correlations, and interference changes outcome probabilities through amplitude addition. Quantum gates are reversible unitary transformations, while measurement converts quantum information into classical data. Circuits combine these operations and are executed repeatedly to estimate distributions.

The dimension of an n-qubit state grows as 2ⁿ, but this does not provide direct access to 2ⁿ classical answers. Useful advantage depends on algorithmic structure, interference, input preparation, output extraction, error rates and comparison with strong classical methods. Complexity classes such as P, BPP and BQP help describe theoretical capability, but practical speedup requires fair resource and accuracy analysis.

Quantum computing differs from classical HPC, quantum communication and quantum sensing in purpose and information-processing model, although hybrid systems are common. Applications include simulation, optimisation, cryptography and scientific modelling. Current systems face noise, limited connectivity, sampling cost and substantial error-correction overhead. Responsible professionals must benchmark honestly, protect data, plan for post-quantum security and distinguish demonstrated performance from future potential.

Key Terms

Qubit: A two-level quantum information unit.

Computational basis: The basis consisting of |0⟩ and |1⟩ for one qubit.

Probability amplitude: A complex coefficient whose squared magnitude gives measurement probability.

Superposition: A coherent linear combination of basis states.

Relative phase: The phase difference between amplitudes that can affect interference.

Entanglement: A multi-system quantum correlation that cannot be represented as independent states.

Interference: Addition of amplitudes that can amplify or suppress outcomes.

Measurement: The process that produces a classical result from a quantum state.

Unitary operator: A reversible transformation satisfying U†U = I.

Quantum gate: An elementary operation, usually unitary, applied to qubits.

Quantum circuit: An ordered arrangement of quantum gates, measurements and qubit wires.

Shot: One execution of a circuit producing one measured outcome.

Circuit depth: The number of sequential gate layers.

BQP: Problems solvable by a quantum computer in polynomial time with bounded error.

Quantum advantage: Demonstrated superior performance for a defined task and resource model.

Quantum error correction: Encoding quantum information to detect and correct errors.

Error mitigation: Techniques that reduce the effect of errors in estimated results.

Quantum communication: Use of quantum states and classical channels for communication tasks.

Quantum sensing: Use of quantum effects to measure physical quantities.

Post-quantum cryptography: Cryptography designed to resist attacks from quantum and classical computers.

Review Questions

5 Short Answer Questions

1. What distinguishes a qubit superposition from a classical probabilistic mixture?
2. State the normalisation condition for a single-qubit pure state.
3. What does a CNOT gate do when its control qubit is 1?
4. Why must quantum circuits usually be executed many times?
5. Name two differences between quantum computing and quantum sensing.

5 Long Answer Questions

1. Explain superposition, entanglement and interference, and show how they cooperate in a quantum circuit.
2. Describe the H–Z–H example and explain why the final measurement is deterministic in the ideal case.
3. Discuss the meaning and limitations of the 2ⁿ-dimensional state space of n qubits.
4. Compare quantum computing with classical HPC, quantum communication and quantum sensing.
5. Explain the main sources of practical limitation in noisy quantum processors.

3 Analytical / Application Questions

1. A team claims quantum advantage because its circuit uses fewer logical steps than an HPC program. Design a fair evaluation plan, including baseline, input-loading cost, output precision and uncertainty.
2. For a Bell-state experiment, explain what results are expected when both qubits are measured in the computational basis, and why those results do not enable faster-than-light communication.
3. A company stores encrypted health records for decades. Analyse why quantum computing affects its security planning and recommend a responsible migration approach.

Multiple Choice Questions with Answers

Question 1

Which expression correctly represents a normalised pure single-qubit state?

A. |ψ⟩ = α|0⟩ + β|1⟩ with |α|² + |β|² = 1

B. |ψ⟩ = α|0⟩ + β|1⟩ with |α| + |β| = 2

C. |ψ⟩ = α|0⟩ + β|1⟩ with α and β required to be real

D. |ψ⟩ = |0⟩ + |1⟩ without any normalisation condition

Correct Answer: A. |ψ⟩ = α|0⟩ + β|1⟩ with |α|² + |β|² = 1

Question 2

What is the main role of relative phase in a quantum circuit?

A. It changes the physical size of a qubit

B. It can affect interference and therefore measurement probabilities

C. It guarantees that measurement outcomes are deterministic

D. It converts a qubit into a classical bit before measurement

Correct Answer: B. It can affect interference and therefore measurement probabilities

Question 3

Starting with |00⟩, applying H to the first qubit and then a CNOT produces which state?

A. |00⟩

B. |01⟩

C. (|00⟩ + |11⟩)/√2

D. (|01⟩ + |10⟩)/√2

Correct Answer: C. (|00⟩ + |11⟩)/√2

Question 4

Why are repeated shots used in quantum computing?

A. To make a unitary gate non-reversible

B. To estimate the probability distribution of measurement outcomes

C. To increase the number of physical qubits automatically

D. To eliminate all systematic hardware errors

Correct Answer: B. To estimate the probability distribution of measurement outcomes

Question 5

Which condition characterises an ideal unitary gate U?

A. U + U† = 0

B. U†U = I

C. U² = I for every gate

D. U must always be diagonal

Correct Answer: B. U†U = I

Question 6

Which statement about quantum advantage is most accurate?

A. A larger qubit count alone proves practical advantage

B. Any quantum algorithm is faster than every classical algorithm

C. Advantage must be defined relative to a task, baseline, resources and accuracy

D. Quantum advantage means that measurement returns every amplitude

Correct Answer: C. Advantage must be defined relative to a task, baseline, resources and accuracy

Question 7

Which field primarily uses quantum systems to measure physical quantities with high sensitivity?

A. Quantum sensing

B. Classical compilation

C. Quantum error correction only

D. Classical database indexing

Correct Answer: A. Quantum sensing

Question 8

For an ideal computational-basis measurement of α|0⟩ + β|1⟩, what is the probability of outcome 1?

A. |α|²

B. |β|²

C. α + β

D. 1 − |α| − |β|

Correct Answer: B. |β|²

Question 9

What is a key limitation of near-term noisy quantum processors?

A. They can directly read all 2ⁿ amplitudes of an n-qubit state

B. They have no need for classical control systems

C. Noise, connectivity constraints and sampling costs limit useful circuit size

D. They cannot perform any reversible operation

Correct Answer: C. Noise, connectivity constraints and sampling costs limit useful circuit size

Question 10

Which statement about entanglement is correct?

A. It permits faster-than-light transmission of controllable messages

B. It is always equivalent to two independent random bits

C. It creates correlations that cannot generally be represented by independent local states

D. It occurs only after measurement has destroyed the quantum state

Correct Answer: C. It creates correlations that cannot generally be represented by independent local states