R. H. Tütüncü, K. C. Toh, and M. J. Todd present SDPT3, a MATLAB implementation of infeasible primal-dual path-following algorithms for solving semidefinite-quadratic-linear programs (SQLPs). The software uses Mehrotra-type predictor-corrector variants of interior-point methods and two search directions: HKM and NT. It solves SQLPs with constraints involving semidefinite, quadratic, and linear cones. SDPT3 can handle problems with semidefinite, second-order, and linear blocks, and it solves both the primal and dual problems. The software is compatible with MATLAB versions 5.3 and 6.0 and is available online. Version 3.0 improves upon previous releases by being faster, especially for large sparse problems, and can directly solve problems with second-order cone constraints. The algorithm computes search directions using Schur complement equations and incorporates the Sherman-Morrison-Woodbury formula for efficient computation. The software also includes an initial iterate strategy that ensures the initial point is of appropriate magnitude for the problem. Computational experiments on SDPLIB and DIMACS Challenge problems demonstrate the effectiveness of SDPT3 in solving SQLPs. The software is flexible, allowing different search directions and options for step-length determination. It is designed to handle large-scale problems efficiently, with a focus on sparsity and computational performance. The implementation details include a cell array representation for problem data, which optimizes memory usage and computational efficiency. The software is robust and can detect infeasibility in cases where the primal or dual problems are infeasible. Overall, SDPT3 is a powerful tool for solving SQLPs with a wide range of applications in optimization.R. H. Tütüncü, K. C. Toh, and M. J. Todd present SDPT3, a MATLAB implementation of infeasible primal-dual path-following algorithms for solving semidefinite-quadratic-linear programs (SQLPs). The software uses Mehrotra-type predictor-corrector variants of interior-point methods and two search directions: HKM and NT. It solves SQLPs with constraints involving semidefinite, quadratic, and linear cones. SDPT3 can handle problems with semidefinite, second-order, and linear blocks, and it solves both the primal and dual problems. The software is compatible with MATLAB versions 5.3 and 6.0 and is available online. Version 3.0 improves upon previous releases by being faster, especially for large sparse problems, and can directly solve problems with second-order cone constraints. The algorithm computes search directions using Schur complement equations and incorporates the Sherman-Morrison-Woodbury formula for efficient computation. The software also includes an initial iterate strategy that ensures the initial point is of appropriate magnitude for the problem. Computational experiments on SDPLIB and DIMACS Challenge problems demonstrate the effectiveness of SDPT3 in solving SQLPs. The software is flexible, allowing different search directions and options for step-length determination. It is designed to handle large-scale problems efficiently, with a focus on sparsity and computational performance. The implementation details include a cell array representation for problem data, which optimizes memory usage and computational efficiency. The software is robust and can detect infeasibility in cases where the primal or dual problems are infeasible. Overall, SDPT3 is a powerful tool for solving SQLPs with a wide range of applications in optimization.