If you only have these two variables, and want to find out the age difference, then the age is the independent variable, since you are using their ages to find out how their preferences vary. So you are trying to predict their preference in computers based on their ages. Basically, you are asking, can I use their ages to predict their preference? See it more as response variable and explanatory variable. Their ages is trying to 'explain' their 'response' to the product.

Preference ~ Age

To make it easier to grasp you could also add 'money' as another explanatory variable. People might base their preference of computer based both on their age AND how much money they can spend.

Preference ~ Age + Money

Another way of seeing it is that you can change their preference based on their age, but you can't change their ages based on preference.

What are the dependent and Independent variable (s)??

Merely based on your research question i.e. " Is there a difference in average age among computer buyers based on their preference for Mac or PCs?" - sounds like you can use analysis of variance e.g. dependent t-test rather than correlational analysis using dependent & independent variables.

If your research conceptual framework / research model is based on correlational study with multiple independent variables (IV) & a dependent variable (DV) - may be DV can be Mac / PC preference, IV should be more e.g. social economic status, previous computer brand experience, friends pressure, apps to be run etc. beside age as the IV. Because there can be other IVs carry more weight in influencing a buyer's decision to purchase Mac over PC instead of merely based on age.