Quantum Computing: Physics Behind the Revolution
Introduction
Quantum computing harnesses the principles of quantum mechanics to process information in ways that classical computers cannot match. Instead of classical bits that are either zero or one, quantum computers use qubits that can exist in superpositions of both states simultaneously. This ability, combined with quantum entanglement and interference, allows quantum computers to solve certain problems exponentially faster than any classical computer.
The physics behind quantum computing draws directly on the quantum mechanical principles developed over the past century. Superposition, entanglement, measurement, and decoherence are not abstract concepts but practical considerations that determine whether a quantum computer will work. Understanding the physics is essential for appreciating both the power and the challenges of quantum computing.
Qubits: The Quantum Bit
Physical Realizations of Qubits
A qubit is any quantum system with two distinguishable states. Many physical systems can serve as qubits: the spin of an electron, the polarization of a photon, the energy levels of an atom or ion, the current direction in a superconducting loop, or the charge state of a semiconductor quantum dot.
Each qubit implementation has its own advantages and challenges. Trapped ion qubits have long coherence times — the duration over which quantum information is preserved — but are difficult to scale. Superconducting qubits are easier to fabricate and control using existing semiconductor manufacturing techniques but have shorter coherence times. Photonic qubits are excellent for communication but challenging for computation because photons do not interact strongly with each other.
Representing Information
A classical bit can be represented as a point at either end of a line segment. A qubit can be represented as a point anywhere on the surface of a sphere — the Bloch sphere. The north pole corresponds to zero, the south pole to one, and every other point corresponds to a superposition of zero and one.
The qubit’s ability to exist anywhere on the Bloch sphere gives quantum computing its power. But it is not just superposition that matters — entanglement between multiple qubits is essential for quantum speedup. A system of n qubits can exist in a superposition of all possible classical states simultaneously, requiring classical computers to track exponentially many amplitudes.
Quantum Gates and Circuits
Single-Qubit Gates
Quantum gates are operations that transform qubit states. They are represented mathematically by unitary matrices — matrices whose inverse equals their conjugate transpose. Single-qubit gates rotate the qubit state on the Bloch sphere. The Pauli gates rotate by 180 degrees around the X, Y, or Z axis. The Hadamard gate creates superpositions, transforming zero into an equal superposition of zero and one.
These gates are analogous to classical logic gates but with important differences. Quantum gates are reversible — information is not lost during computation. They act on probability amplitudes, not just probabilities, enabling quantum interference that can cancel incorrect answers and amplify correct ones.
Two-Qubit Gates and Entanglement
Two-qubit gates are essential for creating entanglement. The controlled-NOT gate flips the target qubit if and only if the control qubit is one. Applied to a superposition state, the CNOT gate creates entangled states that have no classical analog.
A universal set of quantum gates — sufficient to implement any quantum computation — consists of single-qubit rotations and a two-qubit entangling gate. Different qubit technologies implement these gates differently, using laser pulses for trapped ions, microwave pulses for superconducting qubits, or nonlinear optical effects for photonic qubits.
Quantum Algorithms
Shor’s Algorithm
Peter Shor’s 1994 algorithm for factoring large integers demonstrated the potential of quantum computing. Factoring an n-digit number using the best classical algorithm requires time that grows exponentially with n. Shor’s algorithm requires time that grows only polynomially, meaning that a sufficiently large quantum computer could break RSA encryption, which relies on the difficulty of factoring.
Shor’s algorithm uses quantum Fourier transform to find the period of a modular exponential function. The quantum Fourier transform is exponentially faster than the classical fast Fourier transform because it exploits quantum superposition and interference. The algorithm’s existence motivated massive investment in quantum computing research.
Grover’s Algorithm
Lov Grover’s algorithm solves unstructured search problems. Given a database of N items, finding a specific item classically requires checking items one by one — N/2 checks on average. Grover’s algorithm finds the item in roughly square root of N checks, a quadratic speedup.
While less dramatic than Shor’s exponential speedup, Grover’s algorithm is widely applicable. It can accelerate any problem that can be phrased as searching for a solution among many candidates, including optimization, machine learning, and cryptography. The quantum principles underlying these algorithms are the same superposition, entanglement, and interference that power all quantum computation.
Quantum Error Correction
The Problem of Decoherence
Quantum systems are extremely sensitive to their environment. Interactions with stray electromagnetic fields, thermal fluctuations, and material defects cause decoherence — the loss of quantum information. Decoherence limits the time available for computation and is the primary obstacle to building practical quantum computers.
Decoherence cannot be eliminated entirely, but it can be managed through quantum error correction. Unlike classical error correction, which simply copies bits, quantum error correction must work without measuring the quantum state directly, because measurement would collapse the superposition.
Topological Quantum Computing
An alternative approach to error correction is topological quantum computing, which encodes quantum information in nonlocal properties of physical systems that are inherently protected from decoherence. Anyons — quasiparticles that exist only in two-dimensional systems — have braiding statistics that can be used to perform quantum gates. When anyons are exchanged, their worldlines braid around each other, and the resulting transformation depends only on the topology of the braids, not on the details of the path.
Topological quantum computing is inherently fault-tolerant because local perturbations cannot change the topological properties of the braids. The main challenge is finding or engineering materials that support non-Abelian anyons. The fractional quantum Hall effect and certain topological superconductors are promising platforms. Microsoft has invested heavily in topological qubits based on Majorana zero modes, though experimental progress has been slow.
Surface Codes
The surface code is the most promising approach to quantum error correction. It arranges qubits on a two-dimensional grid and uses parity measurements to detect and correct errors without disturbing the stored quantum information. Surface codes require many physical qubits to encode a single logical qubit — current estimates suggest thousands of physical qubits per logical qubit.
The threshold theorem states that if the physical error rate is below a certain threshold, quantum error correction can reduce the error rate arbitrarily by using more qubits. Achieving error rates below this threshold — currently about one percent for surface codes — is a major engineering goal.
Quantum Supremacy and Benchmarking
Quantum supremacy is the milestone at which a quantum computer performs a calculation that is infeasible for any classical computer. Google claimed this milestone in 2019 with their Sycamore processor, performing a random circuit sampling task in 200 seconds that they estimated would take a classical supercomputer 10,000 years. IBM disputed the claim, showing that improved classical algorithms could reduce the estimated classical time.
Random circuit sampling is not a useful task in itself, but it serves as a benchmark for quantum processors. Other benchmarks include quantum volume, which measures the effective size of the largest random circuit that a quantum computer can successfully execute, and cross-entropy benchmarking, which compares the output distribution of a quantum processor to the ideal distribution predicted by classical simulation.
Achieving quantum advantage for a practically useful problem remains an open goal. Potential candidates include quantum simulation of chemical systems, optimization problems, and machine learning tasks. The demonstration of a useful quantum advantage would mark a transition from academic curiosity to practical technology.
Current Challenges and Prospects
Scaling
Current quantum processors have tens to hundreds of qubits. Useful quantum computers will require thousands or millions of logical qubits, each encoded in many physical qubits. Scaling requires improvements in qubit fabrication, control electronics, cryogenics, and error correction.
Different qubit technologies face different scaling challenges. Superconducting qubits require dilution refrigerators and microwave control lines that become increasingly complex as the qubit count grows. Trapped ion qubits require laser systems and vacuum chambers. Photonic qubits require low-loss optical components and efficient photon sources and detectors.
Near-Term Applications
Before fault-tolerant quantum computers are available, noisy intermediate-scale quantum devices may find useful applications. Quantum simulation — using controllable quantum systems to study other quantum systems — is a promising near-term application. Quantum chemistry simulations could accelerate the discovery of new materials, catalysts, and pharmaceuticals.
Quantum sensing — using quantum effects to make ultra-precise measurements — is already producing practical devices. Quantum magnetometers, atomic clocks, and gravitational sensors outperform their classical counterparts. The physics of quantum tunneling and superposition underlies many of these sensing technologies.
What is quantum supremacy? Quantum supremacy is the demonstration that a quantum computer can solve a problem that is infeasible for classical computers. Google claimed quantum supremacy in 2019 with a specialized computation on a 53-qubit processor.
Will quantum computers replace classical computers? No. Quantum computers excel at specific problems like factoring and quantum simulation but are not faster for general-purpose computing. Classical computers will remain essential for most applications.
When will practical quantum computers exist? Predictions vary widely. Fault-tolerant quantum computers with millions of logical qubits are likely at least a decade away. Near-term quantum devices may find specialized applications within a few years.