the two are related. and the explanation is a little backwards. having determinant zero is not necessary for a solution to exist but rather it is sufficient.
(this discussion only applies to maps between spaces of the same dimension.)
if the determinant is non zero, then an n dimensional block is transformed into another n dimensional block, i.e. the dimension of the image space is the same as that of the source space.
It follows that the image space is equal to the entire target space, and hence that every equation has a solution. On the other hand even if a linear map from n space to n space lowers dimension, so that the image is a proper subspace of the target, some equations will still have solutions, but not all.