|
Message-ID: <CAKGDhHUx=DrM0s=i1pUQahDV4oknRSx1Aw1i0c-TO+yHiTUgUw@mail.gmail.com> Date: Tue, 1 Sep 2015 19:34:42 +0200 From: Agnieszka Bielec <bielecagnieszka8@...il.com> To: john-dev@...ts.openwall.com Subject: Re: PHC: Argon2 on GPU 2015-09-01 19:06 GMT+02:00 Solar Designer <solar@...nwall.com>: > Agnieszka, > > On Tue, Sep 01, 2015 at 06:39:06PM +0200, Agnieszka Bielec wrote: >> 2015-08-31 10:53 GMT+02:00 Solar Designer <solar@...nwall.com>: >> > I tried to improve this with a script: >> > >> > [solar@...er run]$ cat try5.sh >> > #!/bin/sh >> > >> > for LWS in 12 16 18 24 32 48 64 96 128 192; do >> > for m in 3 5 7 9 11 13 15 17 19 21; do >> > GWS=$[$LWS * $m] >> > while [ $GWS -lt 1536 ]; do >> > GWS=$[$GWS * 2] >> > done >> > echo $LWS $GWS >> > export LWS GWS >> > ./john -test -format=argon2d-opencl -dev=5 -v=4 >> > done >> > done >> >> Did you checked argon2i by using this script? > > No. > >> I modified argon2i to use ulong4 only when accessing to global memory >> and is slightly faster for specific GWS and LWS. don't know if it's >> faster overall but speed of argon 2i is better than previous record >> >> previously - 6113 >> now - 6301 > > What GPU are these on? TITAN X [a@...er run]$ LWS=128 GWS=4096 ./john --test --format=argon2i-opencl --v=4 --dev=6 Benchmarking: argon2i-opencl [Blake2 OpenCL]... memory per hash : 1.50 MB Device 6: GeForce GTX TITAN X Options used: -I ./kernels -cl-mad-enable -cl-nv-verbose -D__GPU__ -DDEVICE_INFO=131090 -DDEV_VER_MAJOR=352 -DDEV_VER_MINOR=21 -D_OPENCL_COMPILER -DBINARY_SIZE=256 -DSALT_SIZE=64 -DPLAINTEXT_LENGTH=125 Local worksize (LWS) 128, global worksize (GWS) 4096 using different password for benchmarking DONE Speed for cost 1 (t) of 3, cost 2 (m) of 1536, cost 3 (l) of 1 Many salts: 6301 c/s real, 6301 c/s virtual Only one salt: 6301 c/s real, 6301 c/s virtual
Powered by blists - more mailing lists
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.