<p>Existing neural dynamic solvers excel at handling equality-constrained quadratic programming problems, but tackling inequality constraints is crucial for practical applications such as robotic joint limits. When addressing time-related problems, existing methods are mostly developed for continuous-form framework, and there is a lack of independent and effective solution within discrete-form framework. This article builds on the direct discretization approach and integrates existing techniques (penalty function and nonlinear complementarity problem function) to address inequality constraints, proposing two discrete-form recurrent neural dynamic solvers, i.e., the discrete-form recurrent neural dynamic solver with penalty function and that with nonlinear complementarity problem function. These solvers efficiently solve the time-dependent multi-constrained quadratic programming problem in discrete-form framework. Subsequently, the convergence of the above solvers is analyzed theoretically. Furthermore, numerical simulations and comparative studies validate the effectiveness and superiority of the two solvers. Finally, both neural dynamic solvers are applied to the tracking control task of a 6-degree-of-freedom planar robotic manipulator to verify their applicability in practical scenarios.</p>