I have to admit, it seems like single mode would only make a real difference for slow hashes for which each target password itself has a known per-password plaintext component (such as the Myspace leak).
That being said, something in that patpro post gave me pause: "If your file is 60M hashes long, then your GPU will only try 16.6 candidates per second (1000/60). It's very slow and inefficient." That's not the way it works, as I understand it. Is this a CPU vs GPU argument, or a JtR vs hashcat argument?
I'm not sure that I'm following why "hashcat on an Nvidia GTX 1080 is more than 200 times slower than John the Ripper on a single CPU core on this very particular job." Aren't the target hashes described in the article nothing more than salted SHA1? [Edit: OK, I think I understand this now ... for this very specific use case, I assume that optimizing it to work on GPU would probably not be worth the trouble, and that JtR only does this on CPU?]
I sense a gap in my understanding that must be closed.
[Edit: it might help if I'd read the CynosurePrime post. Taking a look]
[Deleted very incorrect analysis of speeds]
Yeah, I think I've been spoiled and have spent too much time working with unsalted hashes. The speed difference might indeed be significant - but I'll have to defer to people with more experience than me on this one, before I dig this hole any deeper.
That being said, something in that patpro post gave me pause: "If your file is 60M hashes long, then your GPU will only try 16.6 candidates per second (1000/60). It's very slow and inefficient." That's not the way it works, as I understand it. Is this a CPU vs GPU argument, or a JtR vs hashcat argument?
I'm not sure that I'm following why "hashcat on an Nvidia GTX 1080 is more than 200 times slower than John the Ripper on a single CPU core on this very particular job." Aren't the target hashes described in the article nothing more than salted SHA1? [Edit: OK, I think I understand this now ... for this very specific use case, I assume that optimizing it to work on GPU would probably not be worth the trouble, and that JtR only does this on CPU?]
I sense a gap in my understanding that must be closed.
[Edit: it might help if I'd read the CynosurePrime post. Taking a look]
[Deleted very incorrect analysis of speeds]
Yeah, I think I've been spoiled and have spent too much time working with unsalted hashes.
The speed difference might indeed be significant - but I'll have to defer to people with more experience than me on this one, before I dig this hole any deeper.
~