This paper presents an efficient and numerically stable dual algorithm for solving strictly convex (positive definite) quadratic programming problems. The algorithm leverages the unconstrained minimum of the objective function as a starting point, utilizing Cholesky and QR factorizations for implementation. The performance of the dual algorithm is compared with primal algorithms using randomly generated test problems and quadratic programs derived from nonlinear programming problems solved by successive quadratic programming methods. Computational results show that the dual algorithm outperforms primal algorithms when a primal feasible point is not readily available. The algorithm is also theoretically compared with modified simplex-type dual methods and illustrated with a numerical example. The paper highlights the efficiency and robustness of the dual approach, particularly in scenarios where a feasible point is challenging to obtain.This paper presents an efficient and numerically stable dual algorithm for solving strictly convex (positive definite) quadratic programming problems. The algorithm leverages the unconstrained minimum of the objective function as a starting point, utilizing Cholesky and QR factorizations for implementation. The performance of the dual algorithm is compared with primal algorithms using randomly generated test problems and quadratic programs derived from nonlinear programming problems solved by successive quadratic programming methods. Computational results show that the dual algorithm outperforms primal algorithms when a primal feasible point is not readily available. The algorithm is also theoretically compared with modified simplex-type dual methods and illustrated with a numerical example. The paper highlights the efficiency and robustness of the dual approach, particularly in scenarios where a feasible point is challenging to obtain.