Hybrid wordlist + mask performance about hashcat

[PenguinKeeper7]

This is a known limitation of Hashcat. With a wordlist that small, Hashcat fails to parallelise properly and therefore it'll only run on probably a single GPU core instead of hundreds or thousands, hence the massive speed issue. It's why this warning is printed:

The wordlist or mask that you are using is too small.
This means that hashcat cannot use the full parallel power of your device(s).
Unless you supply more work, your cracking speed will drop.
For tips on supplying more work, see: https://hashcat.net/faq/morework

Approaching final keyspace - workload adjusted.

This problem can also be seen with these two masks:

> ./hashcat -m 99999 a -a 3 Penguin?b?b?b?b?b?b?b --quiet --speed-only -O
Speed.#1.........:   129.2 MH/s (0.57ms)

> ./hashcat -m 99999 a -a 3 ?b?b?b?b?b?b?bPenguin --quiet --speed-only -O
Speed.#1.........: 60456.8 MH/s (12.33ms)

Fundamentally similar but hundreds of times slower on one mask than the other

For more information about why this happens, consult the "morework" faq link that Hashcat prints, as shown above

[jukuisma]

I have gotten a bit too complaisant with the warnings. Tested a bigger word list and it fixes the performance as you suggested. Closing this issue.

Cheers, Juho