ABSTRACT
An attempt is made to examine the computational aspects of Geometric Programming within a systematic framework. The many numerical algorithms, which have been proposed, and used extensively on practical optimization problems, are considered in relation to developments in general Nonlinear Programming. An effort is made to provide explanations for some of the computational results reported in the literature for specific Polynomial Programming codes and general conclusions on the question of computational efficiency are drawn. The restricted case of “Posynomial“ Programming is considered in detail first, since it forms the core of algorithms used to solve the general Signomial case, which is dealt with subsequently.