+- hashcat Forum (https://hashcat.net/forum)
+-- Forum: Deprecated; Ancient Versions (https://hashcat.net/forum/forum-46.html)
+--- Forum: Very old oclHashcat-plus Support (https://hashcat.net/forum/forum-23.html)
+--- Thread: MD5(unix) performance (/thread-2710.html)
MD5(unix) performance - mrfurious - 10-13-2013
I'm running oclHashcat-plus against some unix md5 hashes from a shadow file (-m 500). I observed a few things:
The larger the list of hashes, the slower the performance. Is this because of salting?
Even if I attack just 4 hashes, I get only about 40,000 H/s on my Radeon 5970 which should be a pretty good GPU. When I run the Example0 script I get almost 2,000 M/s, so I know the GPU itself works fine doing raw hashing. Is my current performance in line with what other people observe when attacking MD5(unix) hashes? What is the primary cause of the several orders of magnitude performance difference between just pure md5 and the unix version? Is the unix version doing many iterations?
Thanks!
RE: MD5(unix) performance - radix - 10-13-2013
salting and iterations are what slow it down
RE: MD5(unix) performance - epixoip - 10-13-2013
the speed reported is the total speed divided by the number of unique salts, because each plaintext candidate needs to be re-hashed with each unique salt.
RE: MD5(unix) performance - atom - 10-15-2013
(10-13-2013, 02:51 PM)mrfurious Wrote: Is the unix version doing many iterations?
right, md5crypt is doing 1000 iterations of md5. but theres also some sort of obfuscation build into it, so the kernels have to do some additional work