Open kanihal opened 6 years ago
i agree. i have lots of cpu's and would appreciate huge leaps in performance on large dirs (mp3, jpg, mp4 directories)
or possibly reducing the accuracy of the hashing with something like https://github.com/MatthewJohn/shamean ? to speed things up.
Blake3 is also a hashing function parallelizzable on multi core CPUs: https://github.com/BLAKE3-team/BLAKE3
I had ~125 GB of data files and it took > 1 day to complete fdupes run since it uses only a single core.
I think it would be very useful if an option of choosing the number of cores to run and parallelize fudpes accordingly.