Optical quantum computing, which uses single photons, linear optical elements, and single photon detectors, became feasible in 2001. Although initially impractical due to high resource requirements, recent advancements have significantly reduced the overhead, making it a serious contender for large-scale quantum computing. Key challenges include high-efficiency single photon sources, low-loss optical circuits, efficient detectors, and low-loss interfacing. Quantum computing promises exponentially faster computation for specific tasks. It requires scalable physical qubits that can be isolated, initialized, measured, and interacted with. Various physical implementations are being explored, including nuclear magnetic resonance, ion traps, cavity quantum electrodynamics, solid-state, and superconducting systems. Photons have emerged as a leading approach due to their low noise and ease of manipulation. Single photons can be encoded in various degrees of freedom, such as polarization, time bin, or path. A single photon qubit can be represented on the Poincaré sphere, and one-qubit gates can be realized using birefringent waveplates. The controlled-NOT (CNOT) gate is a key entangling gate, but its implementation is challenging due to the need for optical non-linearity. However, quantum teleportation and error encoding techniques have enabled non-deterministic CNOT gates, which can be made deterministic through teleportation. Recent developments, including cluster state quantum computing and error encoding techniques, have dramatically reduced resource overhead, making optical quantum computing more attractive. Fault tolerance is crucial, as quantum computers are highly susceptible to noise. The threshold theorem indicates that if noise is below a certain level, arbitrarily long computations can be performed. Recent results suggest that optical quantum computing is feasible if the product of source and detector efficiency exceeds 2/3. Single photon sources, detectors, and circuits are critical components. Current sources rely on non-linear crystals, but solid-state sources show promise. High-efficiency detectors and low-loss optical waveguide circuits are needed for scalability. Nonlinear and hybrid approaches, such as using two-photon absorbers or combining linear optics with nonlinear elements, offer potential advantages. Despite progress, much work remains to realize a large-scale optical quantum computer. The most promising approaches include the circuit model, cluster state model, and hybrid methods. Future developments will depend on advancements in photon sources, detectors, and optical circuits, as well as the practicality of nonlinear optics. Optical quantum computing is a promising route to practical quantum computing, with ongoing research aiming to overcome current challenges.Optical quantum computing, which uses single photons, linear optical elements, and single photon detectors, became feasible in 2001. Although initially impractical due to high resource requirements, recent advancements have significantly reduced the overhead, making it a serious contender for large-scale quantum computing. Key challenges include high-efficiency single photon sources, low-loss optical circuits, efficient detectors, and low-loss interfacing. Quantum computing promises exponentially faster computation for specific tasks. It requires scalable physical qubits that can be isolated, initialized, measured, and interacted with. Various physical implementations are being explored, including nuclear magnetic resonance, ion traps, cavity quantum electrodynamics, solid-state, and superconducting systems. Photons have emerged as a leading approach due to their low noise and ease of manipulation. Single photons can be encoded in various degrees of freedom, such as polarization, time bin, or path. A single photon qubit can be represented on the Poincaré sphere, and one-qubit gates can be realized using birefringent waveplates. The controlled-NOT (CNOT) gate is a key entangling gate, but its implementation is challenging due to the need for optical non-linearity. However, quantum teleportation and error encoding techniques have enabled non-deterministic CNOT gates, which can be made deterministic through teleportation. Recent developments, including cluster state quantum computing and error encoding techniques, have dramatically reduced resource overhead, making optical quantum computing more attractive. Fault tolerance is crucial, as quantum computers are highly susceptible to noise. The threshold theorem indicates that if noise is below a certain level, arbitrarily long computations can be performed. Recent results suggest that optical quantum computing is feasible if the product of source and detector efficiency exceeds 2/3. Single photon sources, detectors, and circuits are critical components. Current sources rely on non-linear crystals, but solid-state sources show promise. High-efficiency detectors and low-loss optical waveguide circuits are needed for scalability. Nonlinear and hybrid approaches, such as using two-photon absorbers or combining linear optics with nonlinear elements, offer potential advantages. Despite progress, much work remains to realize a large-scale optical quantum computer. The most promising approaches include the circuit model, cluster state model, and hybrid methods. Future developments will depend on advancements in photon sources, detectors, and optical circuits, as well as the practicality of nonlinear optics. Optical quantum computing is a promising route to practical quantum computing, with ongoing research aiming to overcome current challenges.