[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20100906222723.GA5421@openwall.com>
Date: Tue, 7 Sep 2010 02:27:23 +0400
From: Solar Designer <solar@...nwall.com>
To: john-users@...ts.openwall.com
Subject: Re: NTLM and OpenCL 1.0
Alain, Erik -
Nice work indeed!
On Mon, Sep 06, 2010 at 05:29:09PM -0400, Erik Winkler wrote:
> OpenCL 64-bit (Note that CPU never exceeds 20% utilization during this test - this is all GPU)
> ./john -test=15 -format:NT
> Benchmarking: NT MD4 [OpenCL 1.0]... DONE
> Raw: 2767K c/s real, 14573K c/s virtual
This might also be 20% of your CPU, without the GPU involved. Notice
how the real c/s rate is roughly 19% of the virtual one, which confirms
this guess. Real is what you actually get; virtual is what you would
presumably get at 100% load of one CPU core (and, from your non-OpenCL
benchmark, you know that you could get even more than that).
Yet it is nice to see that we have "something" working, even if it
performs so poorly at this early stage.
Thanks,
Alexander
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.
