Multiplicative Persistence

I watched numberphilevideo the other day with Matt Parker on multiplicative persistence, and got inspired to give it a try in base ten. Since Im a java-guy I started there with longs (aka 64bit signed integers -> 63bit positives).

I use bruteforce with some educated guesses (that hopefully is valid) to avoid checking all the 263 (more than 1018) numbers (landed on less than 140k candidates).

The number in the video was of course one of the numbers I found, but there seems to be rather few at the top (if one ignore 'the equivalent ones').

PersistenceRaw FormNumber
9[ 11, 7, 0 ]34888999
9[ 1, 2, 12 ]177777779
9[ 6, 6, 5 ]7777788999
9[ 21, 3, 3 ]377788888889
9[ 9, 5, 8 ]37777777788899
9[ 6, 27, 1 ]37889999999999999
10[ 12, 7, 2 ]3778888999
10[ 33, 3, 0 ]3888888888889
11[ 19, 4, 6 ]2777777788888899
11[ 4, 20, 5 ]27777789999999999
PersistenceRaw FormNumber
xkcd 230: hamiltonian (by Randall Munroe)