Page 1 of 2
Project 14167
Posted: Sat Feb 16, 2019 11:03 am
by Calcii
Hello all!
These jobs are so hot!
It is impossible to work with a computer during their folding, Windows core load 99%, Linux - 100%!
And why do these WU's in beta status viewtopic.php?f=66&t=31367 arrive on my advanced key?
Re: Project 14167
Posted: Sat Feb 16, 2019 12:19 pm
by toTOW
The project moved to advanced as per this announcement : viewtopic.php?f=24&t=31395
Which GPU do you have ? We need more details ...
Re: Project 14167
Posted: Sat Feb 16, 2019 6:52 pm
by bruce
Calcii wrote:These jobs are so hot!
All jobs are supposed to use all of the resources they can detect that aren't used by foreground tasks.
If you've overclocked your system, FAH may cause it to crash because FAH is better at using the resources that many traditional overclocking benchmarks.
The CPU resources should be able to run as close to 100% as FAH can achieve since everything else will be at a higher priority and will interrupt its CPU processing.
The most common cause of "screen lag" is folding on a GPU. Pause that slot if you have one and confirm if that's your problem. We can help with that, but you'll need to post the first part of FAH's log.
Re: Project 14167
Posted: Sun Feb 17, 2019 4:41 am
by Calcii
toTOW wrote:The project moved to advanced as per this announcement : viewtopic.php?f=24&t=31395
Which GPU do you have ? We need more details ...
"moved to advanced" missed it, thx
GPU is 1080 Ti
Temperature not my problem(about 54 degrees Celsius), a problem in loading up to 99% of GPU because of what there are friezes on browser and other software
Information For All Crunchers: Raising the memory frequency by 500 MHz reduces TPF by 3-4 seconds on that jobs. Maybe i have not tested such overclocking for a long time, but this was not observed on other jobs
Re: Project 14167
Posted: Sun Feb 17, 2019 5:14 pm
by Trotador
And scoring seems to be lower than other units.
Re: Project 14167
Posted: Sun Feb 17, 2019 5:22 pm
by foldy
If you get lag in browser because of FAH GPU usage then in your browser or other software you can try to disable GPU hardware acceleration.
Re: Project 14167
Posted: Sun Feb 17, 2019 5:58 pm
by toTOW
Trotador wrote:And scoring seems to be lower than other units.
As mentioned in beta thread :
toTOW wrote:This project is showing weird scaling on some GPUs (on 1080 for instance), so let us know if you find abnormal PPD.
But as usual, if you don't provide more details (GPU/OS/Drivers/Expected PPD/Actual PPD), your report is useless ...
Re: Project 14167
Posted: Sun Feb 17, 2019 6:06 pm
by toTOW
Here are my numbers :
1080 - 1873 MHz - Windows - (target 800k PPD)
Avg. Time / Frame : 00:01:30 - 765 711,7 PPD
1070 - 2012 MHz - Windows - (target 620k PPD)
Avg. Time / Frame : 00:02:00 - 698 398,6 PPD
980 - 1330 MHz - Windows - (target 380k PPD)
Avg. Time / Frame : 00:02:35 - 475 748,7 PPD
Re: Project 14167
Posted: Sun Feb 17, 2019 8:52 pm
by Trotador
here some data, two cards in the same host with Ubuntu, average joint daily scoring has decreased around 120Kppd, not a lot but noticeable
card 1 : GTX 1080
63ºC
1900 Mhz
Code: Select all
*********************** Log Started 2019-02-16T16:16:05Z ***********************
17:58:59:WU01:FS01:0x21:*********************** Log Started 2019-02-16T17:58:59Z ***********************
17:58:59:WU01:FS01:0x21:Project: 14167 (Run 13, Clone 9, Gen 22)
17:58:59:WU01:FS01:0x21:Unit: 0x0000001d0002894c5c3d11ed2dd0a12e
17:59:21:WU01:FS01:0x21:Completed 0 out of 1250000 steps (0%)
---
21:16:51:WU01:FS01:0x21:Completed 1250000 out of 1250000 steps (100%)
-------
21:17:12:WU02:FS01:0x21:*********************** Log Started 2019-02-16T21:17:11Z ***********************
21:17:12:WU02:FS01:0x21:Project: 14167 (Run 22, Clone 2, Gen 36)
21:17:12:WU02:FS01:0x21:Unit: 0x000000280002894c5c3d11f81d181491
21:17:20:WU01:FS01:Server responded WORK_ACK (400)
21:17:20:WU01:FS01:Final credit estimate, 96866.00 points
21:17:39:WU02:FS01:0x21:Completed 0 out of 1250000 steps (0%)
------
00:34:59:WU02:FS01:0x21:Completed 1250000 out of 1250000 steps (100%)
---
00:35:19:WU00:FS01:0x21:*********************** Log Started 2019-02-17T00:35:19Z ***********************
00:35:19:WU00:FS01:0x21:Project: 14167 (Run 3, Clone 4, Gen 57)
00:35:19:WU00:FS01:0x21:Unit: 0x0000003e0002894c5c3d11d7381221b2
00:35:19:WU00:FS01:0x21:CPU: 0x00000000000000000000000000000000
00:35:36:WU02:FS01:Server responded WORK_ACK (400)
00:35:46:WU00:FS01:0x21:Completed 0 out of 1250000 steps (0%)
--
03:50:45:WU00:FS01:0x21:Completed 1250000 out of 1250000 steps (100%)
03:51:05:WU01:FS01:0x21:*********************** Log Started 2019-02-17T03:51:05Z ***********************
03:51:05:WU01:FS01:0x21:Project: 14167 (Run 12, Clone 18, Gen 7)
03:51:05:WU01:FS01:0x21:Unit: 0x000000080002894c5c3d11ed24f1cf8d
03:51:05:WU01:FS01:0x21:CPU: 0x00000000000000000000000000000000
03:51:18:WU00:FS01:Server responded WORK_ACK (400)
03:51:18:WU00:FS01:Final credit estimate, 97431.00 points
03:51:28:WU01:FS01:0x21:Completed 0 out of 1250000 steps (0%)
03:51:28:WU01:FS01:0x21:Temperature control disabled. Requirements: single Nvidia GPU, tmax must be < 110 and twait >= 900
03:53:26:WU01:FS01:0x21:Completed 12500 out of 1250000 steps (1%)
03:55:23:WU01:FS01:0x21:Completed 25000 out of 1250000 steps (2%)
07:07:42:WU01:FS01:0x21:Completed 1237500 out of 1250000 steps (99%)
07:10:01:WU02:FS01:0x21:*********************** Log Started 2019-02-17T07:10:00Z ***********************
07:10:01:WU02:FS01:0x21:Project: 14167 (Run 27, Clone 14, Gen 18)
07:10:13:WU01:FS01:Server responded WORK_ACK (400)
07:10:13:WU01:FS01:Final credit estimate, 96674.00 points
07:10:31:WU02:FS01:0x21:Completed 0 out of 1250000 steps (0%)
07:10:31:WU02:FS01:0x21:Temperature control disabled. Requirements: single Nvidia GPU, tmax must be < 110 and twait >= 900
07:12:29:WU02:FS01:0x21:Completed 12500 out of 1250000 steps (1%)
10:27:19:WU02:FS01:0x21:Completed 1237500 out of 1250000 steps (99%)
10:29:16:WU02:FS01:0x21:Completed 1250000 out of 1250000 steps (100%)
10:29:56:WU02:FS01:Server responded WORK_ACK (400)
10:29:56:WU02:FS01:Final credit estimate, 96470.00 points
17:05:59:WU02:FS01:0x21:*********************** Log Started 2019-02-17T17:05:58Z ***********************
17:05:59:WU02:FS01:0x21:Project: 14167 (Run 15, Clone 12, Gen 19)
17:06:16:WU01:FS01:Server responded WORK_ACK (400)
17:06:16:WU01:FS01:Final credit estimate, 96686.00 points
17:06:27:WU02:FS01:0x21:Completed 0 out of 1250000 steps (0%)
20:21:55:WU02:FS01:0x21:Completed 1250000 out of 1250000 steps (100%)
20:22:33:WU02:FS01:Server responded WORK_ACK (400)
20:22:33:WU02:FS01:Final credit estimate, 97283.00 points
card 2: GTX 1080
56ºC
2012 Mhz
Code: Select all
*********************** Log Started 2019-02-16T16:16:05Z ***********************
22:25:32:WU01:FS00:0x21:*********************** Log Started 2019-02-16T22:25:31Z ***********************
22:25:32:WU01:FS00:0x21:Project: 14167 (Run 10, Clone 2, Gen 45)
22:25:32:WU01:FS00:0x21:Unit: 0x000000300002894c5c3d11e867c9bb1d
22:25:55:WU00:FS00:Server responded WORK_ACK (400)
22:25:55:WU00:FS00:Final credit estimate, 99176.00 points
22:26:00:WU01:FS00:0x21:Completed 0 out of 1250000 steps (0%)
--
01:35:58:WU01:FS00:0x21:Completed 1250000 out of 1250000 steps (100%)
01:36:19:WU02:FS00:0x21:*********************** Log Started 2019-02-17T01:36:18Z ***********************
01:36:19:WU02:FS00:0x21:Project: 14167 (Run 1, Clone 11, Gen 20)
01:36:19:WU02:FS00:0x21:Unit: 0x000000180002894c5c3d11d3c9a0dbf5
01:36:27:WU01:FS00:Server responded WORK_ACK (400)
01:36:27:WU01:FS00:Final credit estimate, 98721.00 points
01:36:44:WU02:FS00:0x21:Completed 0 out of 1250000 steps (0%)
--
04:43:59:WU02:FS00:0x21:Completed 1250000 out of 1250000 steps (100%)
04:44:20:WU00:FS00:0x21:*********************** Log Started 2019-02-17T04:44:19Z ***********************
04:44:20:WU00:FS00:0x21:Project: 14167 (Run 31, Clone 6, Gen 47)
04:44:20:WU00:FS00:0x21:Unit: 0x000000330002894c5c3d1205c1e1292e
04:44:27:WU02:FS00:Server responded WORK_ACK (400)
04:44:27:WU02:FS00:Final credit estimate, 99436.00 points
04:44:44:WU00:FS00:0x21:Completed 0 out of 1250000 steps (0%)
---
07:52:12:WU00:FS00:0x21:Completed 1250000 out of 1250000 steps (100%)
07:52:31:WU01:FS00:0x21:*********************** Log Started 2019-02-17T07:52:31Z ***********************
07:52:31:WU01:FS00:0x21:Project: 14167 (Run 29, Clone 18, Gen 9)
07:52:31:WU01:FS00:0x21:Unit: 0x0000000b0002894c5c3d1203e69e9b9b
07:52:38:WU00:FS00:Server responded WORK_ACK (400)
07:52:38:WU00:FS00:Final credit estimate, 99402.00 points
07:52:52:WU01:FS00:0x21:Completed 0 out of 1250000 steps (0%)
---
11:01:01:WU01:FS00:0x21:Completed 1250000 out of 1250000 steps (100%)
11:01:19:WU02:FS00:0x21:*********************** Log Started 2019-02-17T11:01:19Z ***********************
11:01:19:WU02:FS00:0x21:Project: 14167 (Run 29, Clone 12, Gen 24)
11:01:19:WU02:FS00:0x21:Unit: 0x0000001c0002894c5c3d120296174180
11:01:29:WU01:FS00:Server responded WORK_ACK (400)
11:01:29:WU01:FS00:Final credit estimate, 99245.00 points
11:03:32:WU02:FS00:0x21:Completed 12500 out of 1250000 steps (1%)
--
14:10:32:WU02:FS00:0x21:Completed 1250000 out of 1250000 steps (100%)
14:10:52:WU00:FS00:0x21:*********************** Log Started 2019-02-17T14:10:51Z ***********************
14:10:52:WU00:FS00:0x21:Project: 14167 (Run 12, Clone 8, Gen 30)
14:10:52:WU00:FS00:0x21:Unit: 0x000000200002894c5c3d11ec0a695aee
--
14:11:01:WU02:FS00:Server responded WORK_ACK (400)
14:11:01:WU02:FS00:Final credit estimate, 99055.00 points
14:11:18:WU00:FS00:0x21:Completed 0 out of 1250000 steps (0%)
14:13:11:WU00:FS00:0x21:Completed 12500 out of 1250000 steps (1%)
----
17:21:35:WU00:FS00:0x21:Completed 1250000 out of 1250000 steps (100%)
17:21:55:WU01:FS00:FahCore 0x21 started
17:21:56:WU01:FS00:0x21:*********************** Log Started 2019-02-17T17:21:55Z ***********************
17:21:56:WU01:FS00:0x21:Project: 14167 (Run 17, Clone 6, Gen 101)
17:21:56:WU01:FS00:0x21:Unit: 0x0000006a0002894c5c3d11f1c8a6909a
17:22:09:WU00:FS00:Server responded WORK_ACK (400)
17:22:09:WU00:FS00:Final credit estimate, 98624.00 points
--
17:24:14:WU01:FS00:0x21:Completed 12500 out of 1250000 steps (1%)
--
20:29:04:WU01:FS00:0x21:Completed 1250000 out of 1250000 steps (100%)
20:29:31:WU01:FS00:Server responded WORK_ACK (400)
20:29:31:WU01:FS00:Final credit estimate, 99576.00 points
Re: Project 14167
Posted: Mon Feb 18, 2019 10:00 am
by foldy
It takes 03:10 and makes 99k points so it can do 7.7 work units per day which is 762k PPD.
This matches toTOW's numbers for a gtx 1080
Re: Project 14167
Posted: Mon Feb 25, 2019 9:43 pm
by foldy
Can confirm the overall Windows UI lag with this project only. CPU usage is only half a core compared to other projects using a full thread mostly. Maybe there is an improvement of UI experience if the steps done by the work unit would be shorter? So it more often has a short pause in GPU and reloads data from CPU more? Don't know if project owner can configure that.
Re: Project 14167
Posted: Tue Feb 26, 2019 9:07 pm
by toTOW
What you're seeing are the exact limitations of WDDM that prevent small WUs to use efficiently the GPU : WDDM limits the number of commands per second you can send to the GPU to avoid monopolizing resources and to trigger TDR.
If the WU is big enough, each command send to the GPU contains enough data to keep it busy between two commands, so it's more efficient but it also release GPU for UI operation more slowly ... hence the UI lag on such WUs and the higher power usage.
I noticed that there's less UI lag with newer NV drivers on Windows : 418.91 ...
Re: Project 14167
Posted: Thu Feb 28, 2019 12:21 pm
by foldy
If we get more of these projects in the future maybe it is possible to keep the package size between two commands smaller so it is slightly less efficient but UI stays responsive? Or make that an extra option.
Re: Project 14167
Posted: Thu Feb 28, 2019 6:33 pm
by bruce
I'm not sure which package size you're concerned about, but you might check the peak utilization of your paging file and see if you're approaching its limit.
Re: Project 14167
Posted: Thu Feb 28, 2019 9:36 pm
by toTOW
He's talking about OpenCL kernel size. Some GPGPU projects have options in their application for tweaking some values that will move the balance between UI responsiveness and raw compute performances.
I think our older cores (core11) on ATI GPUs had such options ...