I have recently switched employers to one provides educational assistance. This means they will cover the cost of tuition and books at an accredited university. It seems foolish to waste such an opportunity however I have a slight problem in selecting a major for a master's degree. First a quick background: I am currently a computer programmer specializing in distributed computing and I have always been more interested in math. I loaded up on calculus, differential equations, and numerical analysis electives as an undergraduate. I have continued studying these areas informally even though they are not related to my work per se. Now the problem: Ideally I would like to work with physicists, theoretical or experimental, and write the computer programs they need to perform their simulations, data analysis and/or experiments. Do these types of positions even exist? Perhaps they require a PhD? Given the quality of tools and libraries these days I imagine most physicists are able to write their own programs in all but the most extreme cases. (So it is at least possible to work for a tool or library vendor. However, I prefer to be closer to the actual result.) Supposing that such positions were available. Do you think it would be more beneficial to pursue a master's in applied math or computer science with a concentration in numerical techniques? Thanks for your time.