Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 21 Jul 2015 21:00:51 +0200
From: Agnieszka Bielec <>
Subject: Re: PHC: yescrypt on GPU

hi, I discovered right now that benchmarks for yescrypt should be done
using every password different like in pomelo (I don't know why, there
is no coalescing), I could testing it in cracking mode but I wasn't
doing this, and hence is the strange thing with estimating GWS

you should forget about previous benchmarks, sorry that I didn't
discovered this earlier

here is real test (using my modified bench.c)

none@...e ~/Desktop/r/run $ ./john --test --format=yescrypt-opencl
--skip-self-testBenchmarking: yescrypt-opencl [Salsa20/8 OpenCL
(inefficient, development use only)]... Device 0: GeForce GTX 960M
memory per hash : 2.10 MB
Speed for cost 1 (N) of 2048, cost 2 (r) of 8, cost 3 (p) of 11, cost
4 (t) of 0, cost 5 (g) of 0
Many salts:     158 c/s real, 158 c/s virtual
Only one salt:  157 c/s real, 157 c/s virtual

and cracking run:

none@...e ~/Desktop/r/run $ ./john --incremental=digits --min-length=8
--max-length=8 yes --format=yescrypt-opencl
Using default input encoding: UTF-8
Loaded 1 password hash (yescrypt-opencl [Salsa20/8 OpenCL
(inefficient, development use only)])
Device 0: GeForce GTX 960M
memory per hash : 2.10 MB
Press 'q' or Ctrl-C to abort, almost any other key for status
0g 0:00:00:42 0.01% (ETA: 2015-07-29 04:12) 0g/s 155.8p/s 155.8c/s
155.8C/s GPU:72°C util:99% 19910252..11122030
0g 0:00:00:46 0.01% (ETA: 2015-07-29 07:11) 0g/s 155.7p/s 155.7c/s
155.7C/s GPU:72°C util:100% 11122031..20112549

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ