On Fri, 6 Aug 2010 11:28:15 -0700 coderman <coderman@xxxxxxxxx> wrote: > On Fri, Aug 6, 2010 at 6:00 AM, Robert Ransom <rransom.8774@xxxxxxxxx> wrote: > > ... > > | On my 1.5GHz x86-machine, I get about 500k hashes/sec. > > | +---------------------------------------------+ > > | | chars | ~number of tries | ~time @ 500 KH/s | > > Â[snip] > > | | Â Â 6 | Â Â Â32^6 Â= Â1g | Â Â Â Â Â 30 min | > > | | Â Â 7 | Â Â Â32^7 Â= 32g | Â Â Â Â Â Â1 day | > > | | Â Â 8 | Â Â Â32^8 Â= Â1t | Â Â Â Â Â25 days | > > | | Â Â 9 | Â Â Â32^9 Â= 32t | Â Â Â Â2.5 years | > > to throw in my $0.02, this is quite low for any kind of on-demand > computing resource within the same league of difficulty. at $200 you > can get grid/cloud compute time on the order of 200B hashes/hour (160 > cores, 160 G memory) for 3-4 hours. and getting cheaper every day... Good point. Also, the README lists a few optimizations that would make it go even faster (most notably, converting the user-provided regex for the base32-encoded hash to something that the binary hash can be efficiently compared to). Robert Ransom
Attachment:
signature.asc
Description: PGP signature