Regarding gpu-accel & gpu-loops
#1
Hi,

This is an odd question I suppose but I'd like to know if the parameters specified for gpu-accel & gpu-loops set to something higher than what's supported for the support attack in order to gain more performance can result in the result/password/WPA key etc being missed?

For example.

I am currently working on a WPA2 crack with a Radeon 290 that when used with gpu-accel 16 & gpu-loops 16 gives me about 121KH/s.

When used with gpu-accel 80 & gpu-loops 128 I get 179.1KH/s

I've not found anything regarding this question on the forums so I'd just let to get peoples clarification on the matter if I may.

This is the crack running at present,

Session.Name...: virgin-media
Status.........: Running
Input.Mode.....: Mask (?2?2?2?2?2?2?2?2) [8]
Hash.Target....: virginmediaxxxxxxx (00:8e:f2:6b:58:9c <-> 70:14:a6:d9:7b:ba)
Hash.Type......: WPA/WPA2
Time.Started...: Sun Jun 26 19:23:10 2016 (58 secs)
Time.Estimated.: Sun Jul 03 13:19:40 2016 (6 days, 17 hours)
Speed.GPU.#1...:   179.2 kH/s
Recovered......: 0/1 (0.00%) Digests, 0/1 (0.00%) Salts
Progress.......: 6896640000/110075314176 (6.27%)
Rejected.......: 0/6896640000 (0.00%)
Restore.Point..: 287334400/4586471424 (6.26%)
HWMon.GPU.#1...: 100% Util, 70c Temp, 100% Fan

I am running it with options:

oclhashcat32.exe -m 2500 -a3 c:\users\user\Desktop\virginmedia.hccap --custom-charset2=abcdefghjklmnpqrstuvwxyz ?2?2?2?2?2?2?2?2 --gpu-temp-retain 42 --gpu-accel 80 --gpu-loops 128 --session virgin-media  -s 286924800

The only difference (Apart from the increased hash rate), is that the card appears to be working harder and getting hotter than it was set with the 16 / 16 values.


Any help/advice greatly appreciated.
#2
The direct use of -n and -u will no longer be supported with hashcat v3.00 which will automatically tune them for you. You will be able to select a configuration based on the following profiles:

Code:
   | # | Performance | Runtime | Power Consumption | Desktop Impact |
   |---|-------------|---------|-------------------|----------------|
   | 1 | Low         |    2 ms | Low               | Minimal        |
   | 2 | Default     |   12 ms | Economic          | Noticeable     |
   | 3 | High        |   96 ms | High              | Unresponsive   |
   | 4 | Nightmare   |  480 ms | Insane            | Headless       |
#3
(06-29-2016, 12:02 PM)atom Wrote: The direct use of -n and -u will no longer be supported with hashcat v3.00 which will automatically tune them for you. You will be able to select a configuration based on the following profiles:

Code:
   | # | Performance | Runtime | Power Consumption | Desktop Impact |
   |---|-------------|---------|-------------------|----------------|
   | 1 | Low         |    2 ms | Low               | Minimal        |
   | 2 | Default     |   12 ms | Economic          | Noticeable     |
   | 3 | High        |   96 ms | High              | Unresponsive   |
   | 4 | Nightmare   |  480 ms | Insane            | Headless       |

Hi Atom,

Thanks for your response.  Is there a possibility that using values other than what's specified as recommended for WPA2 for example 16 /16 would cause it to skip the correct value or is it a case of finding out what works best to get the best performance out of the GPU ?
#4
(06-29-2016, 12:38 PM)unixm Wrote: Is there a possibility that using values other than what's specified as recommended for WPA2 for example 16 /16 would cause it to skip the correct value or is it a case of finding out what works best to get the best performance out of the GPU ?

No, all values will find the password, if the password is the correct one
#5
(06-29-2016, 03:22 PM)atom Wrote:
(06-29-2016, 12:38 PM)unixm Wrote: Is there a possibility that using values other than what's specified as recommended for WPA2 for example 16 /16 would cause it to skip the correct value or is it a case of finding out what works best to get the best performance out of the GPU ?

No, all values will find the password, if the password is the correct one

Great, thanks for the clarification.