KFA2 1080 ti low utilization
#1
Hello. I have KFA2 1080 ti HoF video card.

That the problem: when I run hashcat, utilization ~ 30%. If in that time I run any 3D game - hashcat utilization ~90%. When game is closed - utils 30%.

If I start any online video, hashcat utilization ~90%. If I pause video, utilization drops till 30% again.

If I start offline video player, like Media Player Classic(even if paused video) - hashcat utilization ~ 90%.

What does it means? How to solve problem?

Sorry, if the same question was before.
Reply
#2
Usually low utilization comes from unsufficient workload for the GPU. Let's concentrate on your hashcat command line. How does it look like?
Reply
#3
Simple: Hashcat64.exe -m 2611 --hex-salt -w 3 --remove vbul_less4.hash -o vbul_less4.found dic\hk_hlm_founds.txt

That happens when i set -w 4 too
Reply
#4
You're missing an attack mode, but that's ok. The dictionary mode is just simply trying out words as password candidates. That's not very effective. Add a rule from the hashcat rule directory like "best64.rule" or "rockyou-30000.rule". Then utilization will be somewhere at 100%.
Reply
#5
Nope. I set rule dictionary best64.rule, utils about 45%. And, if I start MPC (with paused video, as i told before), utils reach 97%.
Here is the MSI Afterburner GPU load graphic(first 90% utils with started MPC, and closed after):

[Image: X93oSetaQD_0BWLH_Zn7Aw.png]


Sometimes, (without any manipulation with video players) utils reach 90% for a short time.
Reply
#6
Forget the video games. The best64.rule already raised the utilization. Try the rockyou-30000.rule or dive.rule.
Reply
#7
I tried to say, that the hashcat rules doesn't matter in this case.
If i run hashcat without rules, utilization about 30%, Brute speed ~800 MH/s.
If in the same time start 3D game or video player(which using GPU for video rendering), utilization reach 90% and Brute speed more 2400 MH/s.

That the problem. And I asked how to solve it without running any application, which using GPU too.
Running hashcat with big rule file that increase the run time in many times - it is not solving the problem.
Reply
#8
I'm just trying to help you. The card should do at least 9000MH/s with that hash kernel - at 100% util.
No idea why it shows higher rates when you start something aside.

But the general problem is, the GPU needs to be fed with enough words from the dictionary. There is a limit for that. The GTX1080Ti is so powerful, you will never reach 100% utilization in pure straight/dictionary mode.

Could you post a status page from when it's running? With and without the video player in the background?
Reply
#9
hashtype md5(md5($pass).$salt)

big dictionary, without rules file, without video player


Code:
Session..........: hashcat
Status...........: Running
Hash.Type........: vBulletin >= v3.8.5
Hash.Target......: for_test.hash
Time.Started.....: Wed Oct 25 22:37:51 2017 (8 secs)
Time.Estimated...: Thu Oct 26 10:17:26 2017 (11 hours, 39 mins)
Guess.Base.......: File (dic\hk_hlm_founds.txt)
Guess.Queue......: 1/1 (100.00%)
Speed.Dev.#2.....:   470.5 MH/s (3.63ms)
Recovered........: 30/134026 (0.02%) Digests, 29/133320 (0.02%) Salts
Recovered/Time...: CUR:N/A,N/A,N/A AVG:0,0,0 (Min,Hour,Day)
Progress.........: 5582905312/19753408594920 (0.03%)
Rejected.........: 2162450400/5582905312 (38.73%)
Restore.Point....: 0/148165381 (0.00%)
Candidates.#2....: ?? -> 502235716829067
HWMon.Dev.#2.....: Temp: 35c Fan: 42% Util: 24% Core:1569MHz Mem:5005MHz Bus:16



big dictionary, without rules file, with video player


Code:
Session..........: hashcat
Status...........: Running
Hash.Type........: vBulletin >= v3.8.5
Hash.Target......: for_test.hash
Time.Started.....: Wed Oct 25 22:37:51 2017 (35 secs)
Time.Estimated...: Thu Oct 26 01:01:10 2017 (2 hours, 22 mins)
Guess.Base.......: File (dic\hk_hlm_founds.txt)
Guess.Queue......: 1/1 (100.00%)
Speed.Dev.#2.....:  2303.9 MH/s (2.89ms)
Recovered........: 30/134026 (0.02%) Digests, 29/133320 (0.02%) Salts
Recovered/Time...: CUR:N/A,N/A,N/A AVG:0,0,0 (Min,Hour,Day)
Progress.........: 16292012000/19753408594920 (0.08%)
Rejected.........: 2162450400/16292012000 (13.27%)
Restore.Point....: 0/148165381 (0.00%)
Candidates.#2....: ?? -> 502235716829067
HWMon.Dev.#2.....: Temp: 42c Fan: 45% Util: 92% Core:1974MHz Mem:5005MHz Bus:16



big dictionary, with small rules file, without video player


Code:
Session..........: hashcat
Status...........: Running
Hash.Type........: vBulletin >= v3.8.5
Hash.Target......: for_test.hash
Time.Started.....: Wed Oct 25 22:38:46 2017 (3 secs)
Time.Estimated...: Mon Oct 30 19:11:07 2017 (4 days, 20 hours)
Guess.Base.......: File (dic\hk_hlm_founds.txt)
Guess.Mod........: Rules (rules\best64.rule)
Guess.Queue......: 1/1 (100.00%)
Speed.Dev.#2.....:  3624.2 MH/s (49.15ms)
Recovered........: 31/134026 (0.02%) Digests, 30/133320 (0.02%) Salts
Recovered/Time...: CUR:N/A,N/A,N/A AVG:19,1169,28069 (Min,Hour,Day)
Progress.........: 174986417760/1521012461808840 (0.01%)
Rejected.........: 166508680800/174986417760 (95.16%)
Restore.Point....: 0/148165381 (0.00%)
Candidates.#2....: ?? -> 502235716829123
HWMon.Dev.#2.....: Temp: 45c Fan: 39% Util: 93% Core:1961MHz Mem:5005MHz Bus:16




small dictionary, with small rules file, without video player


Code:
Session..........: hashcat
Status...........: Running
Hash.Type........: vBulletin >= v3.8.5
Hash.Target......: for_test.hash
Time.Started.....: Wed Oct 25 22:39:57 2017 (6 secs)
Time.Estimated...: Wed Oct 25 23:49:17 2017 (1 hour, 9 mins)
Guess.Base.......: File (dic\_111.txt)
Guess.Mod........: Rules (rules\best64.rule)
Guess.Queue......: 1/1 (100.00%)
Speed.Dev.#2.....:  1887.9 MH/s (4.93ms)
Recovered........: 31/134026 (0.02%) Digests, 30/133320 (0.02%) Salts
Recovered/Time...: CUR:N/A,N/A,N/A AVG:0,0,0 (Min,Hour,Day)
Progress.........: 10658164440/7856181369960 (0.14%)
Rejected.........: 51328200/10658164440 (0.48%)
Restore.Point....: 0/765289 (0.00%)
Candidates.#2....:   -> Shelt62123
HWMon.Dev.#2.....: Temp: 47c Fan: 33% Util: 47% Core:1961MHz Mem:5005MHz Bus:16



small dictionary, with small rules file, with video player


Code:
Session..........: hashcat
Status...........: Running
Hash.Type........: vBulletin >= v3.8.5
Hash.Target......: for_test.hash
Time.Started.....: Wed Oct 25 22:39:57 2017 (27 secs)
Time.Estimated...: Wed Oct 25 23:15:06 2017 (34 mins, 42 secs)
Guess.Base.......: File (dic\_111.txt)
Guess.Mod........: Rules (rules\best64.rule)
Guess.Queue......: 1/1 (100.00%)
Speed.Dev.#2.....:  3756.4 MH/s (4.93ms)
Recovered........: 31/134026 (0.02%) Digests, 30/133320 (0.02%) Salts
Recovered/Time...: CUR:N/A,N/A,N/A AVG:0,0,0 (Min,Hour,Day)
Progress.........: 30929007032/7856181369960 (0.39%)
Rejected.........: 51328200/30929007032 (0.17%)
Restore.Point....: 0/765289 (0.00%)
Candidates.#2....:   -> Shelt62123
HWMon.Dev.#2.....: Temp: 50c Fan: 50% Util: 95% Core:1961MHz Mem:5005MHz Bus:16



small dictionary, with big rules file, without video player


Code:
Session..........: hashcat
Status...........: Running
Hash.Type........: vBulletin >= v3.8.5
Hash.Target......: for_test.hash
Time.Started.....: Wed Oct 25 22:41:10 2017 (3 secs)
Time.Estimated...: Tue Nov 28 22:00:18 2017 (33 days, 23 hours)
Guess.Base.......: File (dic\_111.txt)
Guess.Mod........: Rules (rules\dive.rule)
Guess.Queue......: 1/1 (100.00%)
Speed.Dev.#2.....:  3443.5 MH/s (44.85ms)
Recovered........: 31/134026 (0.02%) Digests, 30/133320 (0.02%) Salts
Recovered/Time...: CUR:N/A,N/A,N/A AVG:0,0,0 (Min,Hour,Day)
Progress.........: 78384044544/10109579054855280 (0.00%)
Rejected.........: 66050727600/78384044544 (84.27%)
Restore.Point....: 0/765289 (0.00%)
Candidates.#2....:   -> Shelt6217$
HWMon.Dev.#2.....: Temp: 49c Fan: 47% Util: 99% Core:1961MHz Mem:5005MHz Bus:16



the same situation with video player and big rules file for md5 hashtype, except run time
Reply
#10
Notice the jump up in utilization when using rules with the big dictionary and the big rules set with the small one. That's as expected.

The interfering video player is odd and interesting. I will try it out myself.

Your big dictionary seems to be huge. Are we talking Terabytes yet? It usually makes more sense to use a dictionary with a size of a few hundred Megabytes (or even smaller) and combine it with rules.
Reply