Just a sub-department at a MA_NG spends $12+ M/year at __A__WS because they're too fucking dainty to do managed colo or rack-and-stack themselves. I'm curious

Just a sub-department at a MA_NG spends $12+ M/year at __A__WS because they're too fucking dainty to do managed colo or rack-and-stack themselves. I'm curious how much N spends nowadays. $100+ M still? Do less computation more efficiently and you won't need to spend obscene amounts on gold-plated rented servers. Capacity plan and manage TCO rather than set money alight.