hashcat Forum

Hi.

What's used to tell if a GPU is being utilized or not?

Because I get the following output:

Code:
Session.Name...: sessions

Status.........: Running

Rules.Type.....: File (rules/oracle)

Input.Mode.....: File dict.dic)

Hash.Target....: File (hash)

Hash.Type......: Oracle 7-10g

Time.Started...: Fri May  3 15:07:16 2013 (2 hours, 10 mins)

Time.Estimated.: Fri May  3 17:25:55 2013 (7 mins, 51 secs)

Speed.GPU.#1...:  1021.1k/s

Speed.GPU.#2...:   952.5k/s

Speed.GPU.#3...:   887.7k/s

Speed.GPU.#4...:   951.7k/s

Speed.GPU.#*...:  3813.1k/s

Recovered......: 0/190 (0.00%) Digests, 0/190 (0.00%) Salts

Progress.......: 3072305320200/3257319317200 (94.32%)

Rejected.......: 4880720/3072305320200 (0.00%)

HWMon.GPU.#1...:  0% Util, 35c Temp, 80% Fan

HWMon.GPU.#2...: 96% Util, 52c Temp, 80% Fan

HWMon.GPU.#3...:  0% Util, 33c Temp, 80% Fan

HWMon.GPU.#4...: 98% Util, 46c Temp, 80% Fan

If Utilization on #1 and #3 is 0%, Speed on #1 and #3 should be zero.. or? If not zero, it should at least not be the same speeds (or higher) than those marking 96+ % utilization...

In this case HWMon.GPU is a more accurate way to tell if a GPU is being utilized. The Speed.GPU that you see there was the last speed calculated/reported for those GPUs.

The work distribution is not always completely equal among all of the GPUs, so some GPUs get more work than others. Most of the time you don't notice, but on long runs like this you are likely to notice that some GPUs will finish their assigned before others.

But would it be feasible to update these last calculated/reported speed so that they show zero like they should be?

yes, please open a trac ticket

splash_

epixoip

mastercracker

atom