The automated search techniques from the field of numerical optimizati
on provide tools that enable optimal design of electronic packages in
general, and solder joints in particular. However, there is considerab
le difficulty in using these procedures for solder joints since the es
timation of fatigue life is computationally very expensive. In this pa
per, global approximation schemes based on designed experiments, linea
r regression models, and artificial neural network models are develope
d to approximate the fatigue life as a function of solder joint design
parameters. Since these approximate surfaces are inexpensive to evalu
ate, their use with the numerical optimization techniques lends to a c
omputationally efficient method for optimizing electronic packages. Th
e developed techniques are demonstrated using the 225 I/O Plastic Ball
Grid Array (PBGA) package, manufactured by Motorola, Inc. An exact op
timization of the solder joints (without approximations) is also carri
ed out and used as a basis for comparing the accuracy and efficiency o
f the developed methods.