The graphical solution procedure generally means you graph the constraint inequalities to identify the feasible solution region, then locate the objective function curve so as to maximize its value. When the objective function is a linear function, maximizing it generally means locating it at the vertex of the feasible region that makes it farthest from the origin.
The solution is
... (A, B) = (100, 50)