Academic grants can work in a lot of ways. It is common for a significant chunk to be taken as overhead by the university (20-40%). This is generally smaller for senior members of the faculty who bring in more grants. The PI (primary investigator, read: dude with a reputation) tends to get 5-10% to run the program, and then another 30-40% goes to salaries for researchers working under them (read: grad students). The rest, on the order of 20%, goes to capital costs like materials, time on expensive machines, or prototypes.
So this guy probably got paid $1-2M directly for the grants over maybe 3-5 years. Note I haven't looked into his specific situation.