Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <49f92df2c432776f38af50e5c3b88d8e@smtp.hushmail.com>
Date: Sat, 26 Jan 2013 11:51:35 +0100
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: Proposed optimizations to pwsafe

Was that with BLOCKS and THREADS tweaked to your specific card, or just the defaults?

magnum


On 26 Jan, 2013, at 11:02 , Brian Wallace <nightstrike9809@...il.com> wrote:

> That makes sense.  It would seem that we can get better performance out of CUDA even in unstable-jumbo.  I tried running it with MPI(just for a sanity test) and it did result in a performance increase, even with the 6 cores accessing the same card.
> 
> On Sat, Jan 26, 2013 at 4:28 AM, magnum <john.magnum@...hmail.com> wrote:
> Bleeding has experimental multi-device support added, and possibly some initial performane regression from that (that we hopefully can fix). 
> magnum
> 
> 
> 
> 
> On 26 Jan, 2013, at 10:12 , Brian Wallace <nightstrike9809@...il.com> wrote:
> 
>> Something I noticed when benchmarking.  For some reason, CUDA runs slower in bleeding than unstable.
>> 
>> On Sat, Jan 26, 2013 at 2:29 AM, Brian Wallace <nightstrike9809@...il.com> wrote:
>> I have some working improvements to pwsafe-cuda.  I don't have the most powerful GPU, but the improvement so far is as follows:
>> 
>> Original:
>> Benchmarking: Password Safe SHA-256 [CUDA]... DONE
>> Raw:    51801 c/s real, 51801 c/s virtual
>> 
>> New:
>> Benchmarking: Password Safe SHA-256 [CUDA]... DONE
>> Raw:    70593 c/s real, 70593 c/s virtual
>> 
>> I do see some other improvements I can make though.
>> 
>> On Fri, Jan 25, 2013 at 8:05 PM, Solar Designer <solar@...nwall.com> wrote:
>> On Fri, Jan 25, 2013 at 05:55:38AM +0100, Lukas Odzioba wrote:
>> > 2013/1/21 Solar Designer <solar@...nwall.com>:
>> > > hashcat's is better, too (354k vs. 501k).  Lukas?
>> >
>> > I have an idea what's need to be done to improve this and some others
>> > sha256 formats.
>> > The problem is that I will not have time for that in the next 3 weeks
>> > because of exams.
>> > I planned for this weekend some other minor/trivial changes that I
>> > think we need before jumbo release, but that's all what I can do at
>> > the moment.
>> > I added this task next to gpu pbkdf2-sha512 to my todo list.
>> 
>> Sounds fine.  Meanwhile:
>> 
>> <gat3way> OK I am in the pwsafe game too. @hashcat you beat me by several K c/s :) http://t.co/CIXK83FT
>> 
>> http://twitter.com/gat3way/status/294968226209726464/photo/1
>> 
>> Alexander
>> 
>> 
> 
> 


Content of type "text/html" skipped

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.