The first step in transformer design optimization is to solve a non-linear optimization task. Here, not only the physical and technological requirements, but the economic aspects are also considered. Large number of optimization algorithms have been developed to solve this task. These methods result the optimal electrical parameters and the shape of the core and winding geometry. Most of them model the windings by their copper filling factors. Therefore the transformer designer’s next task, to find out the detailed winding arrangement, which fits to the optimization results. However, in the case of large power transformers, the calculation of some parameters like: winding gradients, short-circuit stresses etc., needs the knowledge of the exact wire dimensions and winding arrangement. Therefore, an other optimization task should be solved. This paper shows how this sub-problem can be formulated and solved as a generalized geometric program.