Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 27 Apr 2012 02:20:31 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: New RAR OpenCL kernel

On 04/27/2012 12:40 AM, Solar Designer wrote:
> On Thu, Apr 26, 2012 at 11:29:43PM +0200, magnum wrote:
>> But with sane durations (like 3 secs @16K GWS) I now have ~4500 c/s on
>> GTX570/580 as well as HD 7970.
> 
> That's very nice.  There ought to be much room for improvement on the
> 7970, though.

Holy cow. I just replaced an alignment switch statement with a bitflip
macro in sha1_final() - which is only called a mere 17 times in this
256K rounds loop. I just can't believe the result:

Using device 0: Tahiti
Local worksize (LWS) 64, Global worksize (KPC) 65536
Benchmarking: RAR3 (6 characters) [OpenCL]... DONE
Raw:    6725 c/s real, 7245 c/s virtual

...and with fixed length:

Benchmarking: RAR3 (6 characters) [OpenCL]... DONE
Raw:    9033 c/s real, 9899 c/s virtual

That was the least intuitive performance boost I have ever encountered
in my entire life.

magnum

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.