I was actually playing around those files today. I know this will be a typical HN comment and I'm sorry but, if it takes you an hour to load that data into sqlite, you're probably doing something very wrong.

  $ sqlite3 x.db 'create table foo(a,b,c)'

  $ time for year in {1880..2023}; do
      sqlite3 x.db '.import yob'$year'.txt foo --csv';
    done

  real    0m1.359s
  user    0m1.013s
  sys     0m0.323s

  $ sqlite3 x.db 'select count(*) from foo'
  2117219

(Numbers are from Linux/tmpfs/i9-12900k)

Obviously that needs tweaked to store the year in a column, but it shouldn't run for more than a few seconds. Those csv files are the furthest thing from "big data".