|
Message-ID: <AANLkTikUXKZAUkYv1m1crOfKXjdnqeVPYJ-ghHyxY9b-@mail.gmail.com> Date: Fri, 24 Sep 2010 11:56:58 -0700 From: Alain Espinosa <alainesp@...il.com> To: john-users@...ts.openwall.com Subject: Re: NTLM and OpenCL 1.0 On 9/22/10, Charles Weir <cweir@...edu> wrote: > Benchmarking: NT MD4 [OpenCL 1.0]... > OpenCL Platform: <<<Apple>>> and device: <<<GeForce 320M>>> > OpenCL Using direct access to host memory (no transfers) > FAILED (get_hash[0](0)) > > I hope this helps. If you need any additional info please don't hesitate to > ask. The problem with this tests fails it is that i try to access normal RAM from the GPU. Nvidia has a feature called "Zero Copy" in CUDA SDK that made this possible (not all cards support it). In OpenCL do not exists a way to ask for this support, i try a test and error method but as we can see do not work. Maybe this approach works in GPU that use host RAM memory. saludos, alain
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.