Advanced search

Message boards : Number crunching : Very bad display performances

Author Message
jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26226 - Posted: 8 Jul 2012 | 17:48:08 UTC

Hi !
After 2 weeks vacation, I came back and noticed that I received mixed 3.1 and 4.2 WUs on both my machines.

The one under Linux is working fine with much more credits than before.
The one under Win 7 is working but with less credits than before.

On this machine (Win 7) I have a 8 cores CPU(fx 8150) and each GPU uses 0.561 of CPU with a max nr of CPU set to 75%, leaving 2 CPUs for GPUGRID.
World community grid is using only 5 of the 6 remaining cores so 1 CPU seems not be used .
When GPUGRID is running , my display refresh is VERY slow as if the system was slowed down.
I noticed this already sometimes before CUDA 4.2 was used. When suspending GPUs for a while and then reenabling everything was running again file for a while.
But now this is systematically happening : everything is allright when GPUs are not crunching, but slowed down when crunching.
Does someone have the same kind if problem ?

____________
Lubuntu 16.04.1 LTS x64

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26249 - Posted: 9 Jul 2012 | 16:46:24 UTC - in response to Message 26226.

OK !!!
Found that SLI was enabled.
I disabled SLI and everything runs fine again.
Just a warning : if you want also to disable SLI, suspend your crunching or even better wait for all WUs termination (not allowing new work).
I did not do that and got a BSOD !
After reboot the WUs were not in error and continued normally.
____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26497 - Posted: 28 Jul 2012 | 16:53:56 UTC - in response to Message 26249.

OK !!!
Found that SLI was enabled.
I disabled SLI and everything runs fine again.


How/where do you disable SLI??

Thanks, Tom

____________

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26506 - Posted: 28 Jul 2012 | 23:09:42 UTC - in response to Message 26497.
Last modified: 28 Jul 2012 | 23:11:43 UTC

OK !!!
Found that SLI was enabled.
I disabled SLI and everything runs fine again.


How/where do you disable SLI??

Thanks, Tom


Hi Tom !
Go the Nvidia configuration panel.
Go to 3D parameters
here Go to Configure multiple Graphical Processors, Surround and Physx (Note that this a translation from french to english: my panel is in french ;-) )
Then on the right panel, select Disable (or deactivate) multiple graphical processor mode

As a result there must be , at the bottom of the panel, a sentence saying that all graphical processors are operating independently.

Hope this helps...
____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26690 - Posted: 25 Aug 2012 | 9:42:46 UTC - in response to Message 26506.

OK !!!
Found that SLI was enabled.
I disabled SLI and everything runs fine again.


How/where do you disable SLI??

Thanks, Tom


Hi Tom !
Go the Nvidia configuration panel.
Go to 3D parameters
here Go to Configure multiple Graphical Processors, Surround and Physx (Note that this a translation from french to english: my panel is in french ;-) )
Then on the right panel, select Disable (or deactivate) multiple graphical processor mode

As a result there must be , at the bottom of the panel, a sentence saying that all graphical processors are operating independently.

Hope this helps...


Many thanks for responding.

I have Nvidia Control Panel 4.4.370.0. There I clicked "Manage 3D Settings", which gave me two tabs: "Global Settings" and "Program Settings". Here ya go:



I see no reference to SLI there...

I'm really getting desperate over video performance while GPUGRID is running. Most of the day I have to suspend BOINC.
____________

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26696 - Posted: 25 Aug 2012 | 12:07:06 UTC - in response to Message 26690.

Hi Tom !
How many GPU(s) do you have and are your GPU(s) brand and model ?
What is the driver version ?
Does you MOBO has an integrated graphic chip ?

My NVidia Configuration Panel is 4.4.730.0

I suggest you intall latest (at least 301.42, these run fine) driver.
and latest Configuration Panel if not updated by the driver bundle.

The line I am talking about on the left side of the configuration panel is just after the one you choosed i.e "Manage 3D Settings".
I have the same as you have for what you are showing in "Manage 3D Settings".

If you don't have this than you need updating or you don(t have an SLI capable configuration.
____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26698 - Posted: 25 Aug 2012 | 13:05:05 UTC - in response to Message 26696.

Thanks for the fast response, jlhal.

I have a single ASUS GTX 460, OCed to 850/1700/2000. No graphic chip on the MOBO.

I went to nvidia.com looking for a dreiver update. Here's what he said:



Looks like I'm up-to-date.

Here's what I find below "Manage 3D Settings":



So I guess SLI does not apply to me and I must look elsewhere for an answer to my performance problem...
____________

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26699 - Posted: 25 Aug 2012 | 16:03:12 UTC - in response to Message 26698.

You're right, SLI does not apply to you.
I used to run GPUGRID on 2 GTX460 one was a 460SO the other a 460OC . Both OCed from factory and everything was OK (no SLI configured thow)
If your card is not a factory OCed one , I suggest you come back to factory settings and see what is happening...
If this is not the case...I'm wondering...
On normal use the card behaves well, but if you are playing graphical power demanding games, you should invalidate computing while your app is active (cc_config.xml).
____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26701 - Posted: 25 Aug 2012 | 17:18:38 UTC - in response to Message 26699.

If your card is not a factory OCed one , I suggest you come back to factory settings and see what is happening...


Thanks. OK. Did that. We shall see...

But... is there not a setting that will let me select how much - how little - GPU power is devoted to BOINC??

____________

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26702 - Posted: 25 Aug 2012 | 18:52:20 UTC - in response to Message 26701.


But... is there not a setting that will let me select how much - how little - GPU power is devoted to BOINC??


To my knowledge there is actually no paremeter to limit GPU usage in terms of power percentage.
Only the CPU power usage may be limited.

Keep us in touch ...

____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26780 - Posted: 4 Sep 2012 | 7:03:57 UTC

I've been suffering this problem for months (ASUS GTX 460 OCed, Win 7, 8 CPUs)and it's getting so bad now that I've been reluctantly forced to stop GPUGRID processing. I do hope someone will, sometime, come up with a solution and post here.
____________

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26785 - Posted: 4 Sep 2012 | 16:27:54 UTC - in response to Message 26780.

I've been suffering this problem for months (ASUS GTX 460 OCed, Win 7, 8 CPUs)and it's getting so bad now that I've been reluctantly forced to stop GPUGRID processing. I do hope someone will, sometime, come up with a solution and post here.


Are you crunching for other projects and do you have the same problem with them ?
____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26786 - Posted: 4 Sep 2012 | 16:35:11 UTC - in response to Message 26785.

I've been suffering this problem for months (ASUS GTX 460 OCed, Win 7, 8 CPUs)and it's getting so bad now that I've been reluctantly forced to stop GPUGRID processing. I do hope someone will, sometime, come up with a solution and post here.


Are you crunching for other projects and do you have the same problem with them ?


No. I am (was...) a 100% dedicated GPUGRID-er.

____________

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26787 - Posted: 4 Sep 2012 | 17:32:50 UTC

I may, just..., have fixed the problem!

I thought I'd try not running "long runs". Found a parameter in BOINC / Tools / Computing Preferences / processor usage "Use GPU while computer is in use". It was checked. I unchecked it.

I'm now doing "long runs" and will report here what happens...
____________

rypskar
Send message
Joined: 19 Feb 11
Posts: 2
Credit: 220,234,196
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 26791 - Posted: 5 Sep 2012 | 18:52:16 UTC

I am having a similar problem. I have to pause gpugrid when using my PC, but only on the NATAN wus. Other work units can run when I use my PC without any problems.

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27087 - Posted: 17 Oct 2012 | 13:44:54 UTC - in response to Message 26787.

I may, just..., have fixed the problem!

I thought I'd try not running "long runs". Found a parameter in BOINC / Tools / Computing Preferences / processor usage "Use GPU while computer is in use". It was checked. I unchecked it.

I'm now doing "long runs" and will report here what happens...


It made no difference to the poor performance. I've been giving my spare cycles to projects since May 1999, for the last three years to GPUGRID, so it is with great reluctance that I have decided to call it a day.
____________

Profile skgiven
Volunteer moderator
Project tester
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,925,443,810
RAC: 212,063
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27094 - Posted: 19 Oct 2012 | 10:23:28 UTC - in response to Message 27087.

If you select to not use the GPU when the system is in use, and the display performance is still poor, then the problem has nothing to do with running GPUGrid tasks, as these are not running when you are using the system.
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27097 - Posted: 19 Oct 2012 | 17:28:36 UTC - in response to Message 27087.

tomba :
It made no difference to the poor performance. I've been giving my spare cycles to projects since May 1999, for the last three years to GPUGRID, so it is with great reluctance that I have decided to call it a day.

skgiven :
If you select to not use the GPU when the system is in use, and the display performance is still poor, then the problem has nothing to do with running GPUGrid tasks, as these are not running when you are using the system.


Hi tomba, I agree with skgiven,

I did not notice trhat you are connected through HDMI, this is what is shown in your control panel), to a SAMSUNG S22B350 Full HD 22" LED Monitor with VGA and HDMI capabilities.

Can you try connect it through VGA instead , using a DVI to VGA adaptor (normally delivered with the card) and see what happens ?
I don't really know how it works with hdmi on the card side but who knows ?

____________
Lubuntu 16.04.1 LTS x64

Jacob Klein
Send message
Joined: 11 Oct 08
Posts: 1093
Credit: 1,432,402,989
RAC: 1,079,859
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27099 - Posted: 19 Oct 2012 | 20:45:30 UTC
Last modified: 19 Oct 2012 | 20:58:04 UTC

In order for to work around the screen display lag issue....

I have made it a habit to suspend GPU tasks of certain applications, that are known to always generate lag, whenever I sit down at the computer, and then resume them before I leave. I also suspend other GPU tasks if I notice they are causing lag.

Right now, when I sit down at the computer, I always suspend World Community Grid Help Conquer Cancer tasks (since they always cause display lag for me), and I sometimes will suspend a GPUGrid.net task (if its a task that generates display lag that I notice after sitting down to my computer).

That way, while I'm working, my GPU can still crunch the tasks that don't generate lag (like Poem@Home tasks, or most GPUGrid.net tasks), and then when I leave, since I've resumed the other tasks before I leave, the scheduler will eventually complete them too.

That's the only workaround I have. I don't like it, but I cannot otherwise control the lag in some of these GPUGrid.net tasks.

GPUGrid.net Techs:
Should I be keeping track of which GPUGrid.net task types are generating display lag, and reporting them somewhere?

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27101 - Posted: 20 Oct 2012 | 6:45:51 UTC - in response to Message 27094.

If you select to not use the GPU when the system is in use, and the display performance is still poor, then the problem has nothing to do with running GPUGrid tasks, as these are not running when you are using the system.

Sorry, but that's not right, for me. Here are my Tools / Computing Preferences settings:



If I open BOINC in a small window and start Chrome in another (opens 15 tabs in about 15 seconds), BOINC's progress % never waivers. It just keeps on going.
____________

Profile skgiven
Volunteer moderator
Project tester
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,925,443,810
RAC: 212,063
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27103 - Posted: 20 Oct 2012 | 18:21:17 UTC - in response to Message 27101.

So you are saying the tasks keep running, when you start to use the system?
If so, what sort of installation did you do; a Boinc service installation, fresh installation or upgrade?

I would try setting the "While processor usage is less than" to 10%
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27104 - Posted: 20 Oct 2012 | 18:46:40 UTC - in response to Message 27103.

So you are saying the tasks keep running, when you start to use the system?

Yes. Perhaps you missed my comment below the pic: "If I open BOINC in a small window and start Chrome in another (opens 15 tabs in about 15 seconds), BOINC's progress % never waivers. It just keeps on going."

If so, what sort of installation did you do; a Boinc service installation, fresh installation or upgrade?

Having uninstalled BOINC a month ago I downloaded a fresh copy (7.0.28).

I would try setting the "While processor usage is less than" to 10%

Did that.

I just finished a WU - no impact on performance. The current WU, though running 100% of the time, has no impact either.
____________

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27105 - Posted: 20 Oct 2012 | 21:41:32 UTC - in response to Message 27104.

Bonsoir tomba,
your processor is an Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz [Family 6 Model 26 Stepping 5]
(8 processeurs

I notice you let 100% of CPU usage allowed and 100% of processors/threads ?

You have only one GPU so you could set the CPU affinity for GPUGRID WUs in the task manager -> processes -> right click on the process for GPUGRID WU (acemd... for exemple) -> Define affinity -> select the desired cores you let the process use. Unselect others.
By default all cores are allowed..

For exemple let 4 cores for GPUGRID WUs in a total of 8 and observe.

By the way did you read my remark above about HDMI vs VGA ?

Regards
____________
Lubuntu 16.04.1 LTS x64

werdwerdus
Send message
Joined: 15 Apr 10
Posts: 123
Credit: 1,004,473,861
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27107 - Posted: 21 Oct 2012 | 4:02:10 UTC

I would check that you don't have enabled "Leave Applications In Memory While Suspended" in the disk/memory tab. It could be causing problems if the tasks aren't terminated and left in memory, they are still active, somehow.
____________
XtremeSystems.org - #1 Team in GPUGrid

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27109 - Posted: 21 Oct 2012 | 6:41:26 UTC - in response to Message 27107.

I would check that you don't have enabled "Leave Applications In Memory While Suspended" in the disk/memory tab. It could be causing problems if the tasks aren't terminated and left in memory, they are still active, somehow.


No. That option is not checked.
____________

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27110 - Posted: 21 Oct 2012 | 8:11:05 UTC - in response to Message 27105.

Bonsoir tomba,
your processor is an Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz [Family 6 Model 26 Stepping 5]
(8 processeurs

I notice you let 100% of CPU usage allowed and 100% of processors/threads ?

You have only one GPU so you could set the CPU affinity for GPUGRID WUs in the task manager -> processes -> right click on the process for GPUGRID WU (acemd... for exemple) -> Define affinity -> select the desired cores you let the process use. Unselect others.
By default all cores are allowed..

For exemple let 4 cores for GPUGRID WUs in a total of 8 and observe.


OK. Did that, though it seems he does not remember the affinity setting. Not good for unattended operation!

By the way did you read my remark above about HDMI vs VGA ? Regards


Not yet...

Since I restarted processing GPUGRID WUs, two have processed with no performance impact. One of them is uploading now. The third one that's just started has jammed the video brakes on, hard!!




____________

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27111 - Posted: 21 Oct 2012 | 8:20:43 UTC - in response to Message 27105.

By the way did you read my remark above about HDMI vs VGA ?


Oops - my ASUS GTX 460 doesn't have a VGA port!

____________

Profile skgiven
Volunteer moderator
Project tester
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,925,443,810
RAC: 212,063
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27112 - Posted: 21 Oct 2012 | 11:20:37 UTC - in response to Message 27111.

Are you running any CPU tasks on that system?

The recommended power setting is Prefer maximum performance.

If the first thing you do when you start using your system is to load 15 web pages, this probably doesn't give Boinc enough time to suspend the GPU task, which means unloading and writing it to disk (GPU tasks are not kept in memory). If the 15 web pages contain Java or flash then they are going to be resource (HDD and RAM) greedy.

I think this is a general Boinc problem (from legacy designs) which is being exasperated to some extent by task types (size), but largely other programs (browsers), hundreds of operating system updates/patches and of course drivers that are aimed at the most recent cards.

If you ran Boinc from another drive, it might help, but try to just give it 30sec before you open your browser.
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27114 - Posted: 21 Oct 2012 | 16:06:11 UTC - in response to Message 27111.
Last modified: 21 Oct 2012 | 16:11:16 UTC

By the way did you read my remark above about HDMI vs VGA ?


Oops - my ASUS GTX 460 doesn't have a VGA port!



Yes , but your monitor seems to have one.
You must use a DVI to VGA adaptor normally delivered with the card if you bought it new...!
So you will be using the one of the DVI output of the card plugged to the VGA input of the monitor through the adaptor .
You will of course need a VGA cable between the adaptor and the monitor... .
____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27116 - Posted: 21 Oct 2012 | 16:43:53 UTC - in response to Message 27112.

Are you running any CPU tasks on that system?

No.

The recommended power setting is Prefer maximum performance.

OK. Did that.

If the first thing you do when you start using your system is to load 15 web pages, this probably doesn't give Boinc enough time to suspend the GPU task, which means unloading and writing it to disk (GPU tasks are not kept in memory). If the 15 web pages contain Java or flash then they are going to be resource (HDD and RAM) greedy.

But if BOINC is obeying my settings, in 15 seconds it must have time to at least stop the running WU before doing its housekeeping. No sign of it doing that...

I think this is a general Boinc problem (from legacy designs) which is being exasperated to some extent by task types (size), but largely other programs (browsers), hundreds of operating system updates/patches and of course drivers that are aimed at the most recent cards.

If you ran Boinc from another drive, it might help, but try to just give it 30sec before you open your browser.

Did that. No difference.

Seems to me that that some WUs obey my BOINC settings, some don't. Surely that points to a GPUGRID problem??

I'm now keeping a log of which WUs affect performance and which don't.

____________

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 668
Credit: 2,498,095,550
RAC: 12
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27118 - Posted: 22 Oct 2012 | 7:23:49 UTC - in response to Message 27116.
Last modified: 22 Oct 2012 | 7:24:27 UTC

I have had the same problem (NATHAN units only). They can not run on GTX460 while you're using computer without making computer nearly useless.

I would like to say all other units run fine!
____________
Radio Caroline, the world's most famous offshore pirate radio station.
Great music since April 1964. Support Radio Caroline Team -
Radio Caroline

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27121 - Posted: 22 Oct 2012 | 21:36:24 UTC - in response to Message 27118.

I have had the same problem (NATHAN units only). They can not run on GTX460 while you're using computer without making computer nearly useless.


Sorry, but this is not true for me.

I used to have a GTX460SO and a GTX460OC and ALL WUs were running without loss of video performance even when watching a movie and at the exception of playing some games : in this case I simply have forbidden GPU crunching in cc_config.xml when these applications are running .!

____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27125 - Posted: 23 Oct 2012 | 6:03:58 UTC - in response to Message 27121.

... I simply have forbidden GPU crunching in cc_config.xml when these applications are running .!


Can you tell us how to do that? Thanks.

____________

werdwerdus
Send message
Joined: 15 Apr 10
Posts: 123
Credit: 1,004,473,861
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27135 - Posted: 23 Oct 2012 | 20:01:12 UTC - in response to Message 27125.
Last modified: 23 Oct 2012 | 20:01:31 UTC

... I simply have forbidden GPU crunching in cc_config.xml when these applications are running .!


Can you tell us how to do that? Thanks.

<exclusive_gpu_app>important.exe</exclusive_gpu_app>

http://boinc.berkeley.edu/wiki/Cc_config.xml
____________
XtremeSystems.org - #1 Team in GPUGrid

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27145 - Posted: 24 Oct 2012 | 14:45:38 UTC - in response to Message 27116.

I'm now keeping a log of which WUs affect performance and which don't.


In five days:

1. I've had no performance problems with three NOELIAs and one NATHAN
3. One NATHAN gave performance problems
3. Three NOELIAs and one NATHAN stopped with errors
4. Today I aborted four NATHANs for serious performance hits.

Nothing conclusive yet, but it's beginning to look like I need a way to not to get NATHANs...

____________

TJ
Send message
Joined: 26 Jun 09
Posts: 815
Credit: 1,470,348,955
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27149 - Posted: 24 Oct 2012 | 21:48:52 UTC
Last modified: 24 Oct 2012 | 21:49:26 UTC

Jlhal, it could be a driver problem, I have working for several days with a problem of stopping WU's and unresponsive system.
I just downgraded to driver 301.42 and it is working now as I was used it did.
____________
Greetings from TJ

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27150 - Posted: 24 Oct 2012 | 22:01:46 UTC - in response to Message 27149.

Jlhal, it could be a driver problem, I have working for several days with a problem of stopping WU's and unresponsive system.
I just downgraded to driver 301.42 and it is working now as I was used it did.


Hi TJ !
Well, I want to try some day to play again while GPU is crunching and with/out multigpu enabled on my GTX 690
My problem with SLI were with a GTX 590.
All WUs (except for 1 frozen NOELIA I had to cancel after 43h) are running fine on both GPUs at the moment with the latest drivers for Windows 7 and Ubuntu.
____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27190 - Posted: 31 Oct 2012 | 13:28:08 UTC - in response to Message 27145.

I'm now keeping a log of which WUs affect performance and which don't.

Today I aborted eight Natans, one after the other, 'cause each just about froze my PC. I set Gpugrid to 'no new tasks' and will try again tomorrow.


____________

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27349 - Posted: 19 Nov 2012 | 14:58:33 UTC

I found the solution. I did not know that, after unchecking "Use GPU while computer is in use" in BOINC / Tools / Computing Preferences / Processor Usage, I had to activate this preference in BOINC / Activity by checking "Use GPU based on preferences".

Now when I wiggle the mouse in BOINC / Tasks, the GPUGRID WU immediately gives "Waiting to run". Perfect!
____________

TJ
Send message
Joined: 26 Jun 09
Posts: 815
Credit: 1,470,348,955
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27354 - Posted: 19 Nov 2012 | 16:11:38 UTC - in response to Message 27349.

I found the solution. I did not know that, after unchecking "Use GPU while computer is in use" in BOINC / Tools / Computing Preferences / Processor Usage, I had to activate this preference in BOINC / Activity by checking "Use GPU based on preferences".

Now when I wiggle the mouse in BOINC / Tasks, the GPUGRID WU immediately gives "Waiting to run". Perfect!


I had similar problems, with freezing screens or WU'that error out after 15-20 hours running.

First thing I did was downgrading my nVidia driver for the GTX285.
That helped a few weeks, then it happened again.
I found out that Windows was trying to update some software for the graphics card.
I have disabled that and now it runs fine. Same option as you, GPU based on preferences. But my preference is "Run always". And it does without problems so far.
____________
Greetings from TJ

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27522 - Posted: 3 Dec 2012 | 15:36:05 UTC - in response to Message 27349.

I found the solution. I did not know that, after unchecking "Use GPU while computer is in use" in BOINC / Tools / Computing Preferences / Processor Usage, I had to activate this preference in BOINC / Activity by checking "Use GPU based on preferences".

Now when I wiggle the mouse in BOINC / Tasks, the GPUGRID WU immediately gives "Waiting to run". Perfect!


Certainly not perfect! I ran as above for five days. Every day I had two or three PC lock-outs - it was dead - it was big-red-switch time.

So I'm now running as before, and aborting NATHANs that kill my PC performance.

Hello Nathan - are you listening??

____________

Profile Beyond
Avatar
Send message
Joined: 23 Nov 08
Posts: 1073
Credit: 4,619,000,754
RAC: 353,615
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27534 - Posted: 4 Dec 2012 | 4:19:41 UTC - in response to Message 27118.

I have had the same problem (NATHAN units only). They can not run on GTX460 while you're using computer without making computer nearly useless.

I would like to say all other units run fine!

If I may add my 2 cents. I'm currently running 4 GTX460 (768MB) GPUs here. All WUs run perfectly except for NATHAN. The NATHAN WUs run at half speed on all 4 of the above cards. It should be rather easy to up a different queue for the NATHAN WUs by using a unique exe name and allowing us to choose whether to run them via a check box in our project preferences.

Profile Chilean
Avatar
Send message
Joined: 8 Oct 12
Posts: 95
Credit: 330,754,398
RAC: 1,802,419
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwat
Message 27535 - Posted: 4 Dec 2012 | 5:18:52 UTC
Last modified: 4 Dec 2012 | 5:20:42 UTC

The long runs NATHAN make my PC graphics laggy. The short runs don't AT ALL (you can't even tell you are running GPUGRID with the short runs), but given the 150% bonus for sending a long run back, I'm sucking it up (I get around 1.5 credits/second with the long runs versus ~0.7 credits/second with the short runs).

Profile Chilean
Avatar
Send message
Joined: 8 Oct 12
Posts: 95
Credit: 330,754,398
RAC: 1,802,419
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwat
Message 27536 - Posted: 4 Dec 2012 | 7:00:25 UTC - in response to Message 27535.
Last modified: 4 Dec 2012 | 7:01:29 UTC

I just got a "TONI" long run, and OMG what a difference does it make. The graphics on my computer are lag-less. You can't even tell I'm running GPUGRID.

Edit: I just realized it's only using 92% of the GPU vs the 98% the NATHANS use. Still... the difference is amazing (in favor of the TONI).

Profile skgiven
Volunteer moderator
Project tester
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,925,443,810
RAC: 212,063
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27561 - Posted: 5 Dec 2012 | 0:55:09 UTC - in response to Message 27536.

Go TONI
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Profile Beyond
Avatar
Send message
Joined: 23 Nov 08
Posts: 1073
Credit: 4,619,000,754
RAC: 353,615
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27562 - Posted: 5 Dec 2012 | 1:42:37 UTC - in response to Message 27561.
Last modified: 5 Dec 2012 | 1:43:07 UTC

Go TONI

+1

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27571 - Posted: 5 Dec 2012 | 18:57:43 UTC - in response to Message 27536.

I just got a "TONI" long run, and OMG what a difference does it make. The graphics on my computer are lag-less. You can't even tell I'm running GPUGRID.


I'm so glad that it's not just me!! I've just processed a TONI too and all is sweetness and light!

Pretty please, Nathan, take action, ANY action, to stop your WUs killing our PC's performance!! We are dedicated to your cause. Please dedicate yourself to ours!! Thank you.

____________

Profile Beyond
Avatar
Send message
Joined: 23 Nov 08
Posts: 1073
Credit: 4,619,000,754
RAC: 353,615
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27574 - Posted: 5 Dec 2012 | 19:10:33 UTC - in response to Message 27571.

Pretty please, Nathan, take action, ANY action, to stop your WUs killing our PC's performance!! We are dedicated to your cause. Please dedicate yourself to ours!! Thank you.

And please fix them so they won't run at 1/2 speed on some GPUs. Thanks also.

TheFiend
Send message
Joined: 26 Aug 11
Posts: 98
Credit: 1,830,705,479
RAC: 961,552
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27576 - Posted: 5 Dec 2012 | 20:19:01 UTC - in response to Message 27534.

I have had the same problem (NATHAN units only). They can not run on GTX460 while you're using computer without making computer nearly useless.

I would like to say all other units run fine!

If I may add my 2 cents. I'm currently running 4 GTX460 (768MB) GPUs here. All WUs run perfectly except for NATHAN. The NATHAN WUs run at half speed on all 4 of the above cards. It should be rather easy to up a different queue for the NATHAN WUs by using a unique exe name and allowing us to choose whether to run them via a check box in our project preferences.


I was running a GTX460(768MB) and GTX550Ti in my main cruncher and the GTX550 got through Nathans much faster. When I decided to upgrade the 460 was the first to go.

Now running dual GTX670's.

werdwerdus
Send message
Joined: 15 Apr 10
Posts: 123
Credit: 1,004,473,861
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27579 - Posted: 6 Dec 2012 | 1:50:42 UTC

I vote for NOT decreasing the speed of the NATHAN units. IMO they should be coded to run as fast as possible and it seems NATHAN has figured out how to do that. The display lag seems to be an issue with the OS/driver interactions. If it makes your system unusable I would suggest running the standard work units (short) or unchecking "Use GPU while computer is in use."

This is my opinion and I understand that it is not the majority.
____________
XtremeSystems.org - #1 Team in GPUGrid

Profile Beyond
Avatar
Send message
Joined: 23 Nov 08
Posts: 1073
Credit: 4,619,000,754
RAC: 353,615
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27585 - Posted: 6 Dec 2012 | 13:07:24 UTC - in response to Message 27579.

I vote for NOT decreasing the speed of the NATHAN units.

I don't think it has anything to do with the "speed" of the WUs. It seems that the way they're coded they run OK on some cards and not on others. Nathan might want to talk to Toni :-)

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27591 - Posted: 6 Dec 2012 | 17:00:16 UTC

I agree with werdwerdus and sorry if I repeat myself once again...

I have been crunching Nathans WUs with 2 GTX460 (1GB GDDR each if I remember well) on 2 different systems (Win and Ubuntu) and then on one system with both cards together : I never had this problem with Nathan's WUs.

The only problem was later with a GTX590 with SLI enabled (See my 1st post in that was the starting point of this thread).

This has NOTHING to do with Nathan WUs .

Sorry but if you get problems , I' m pretty sure this is a problem with YOUR system.

upgrade or change your GPU or upgrade rour system ! !

____________
Lubuntu 16.04.1 LTS x64

Profile Beyond
Avatar
Send message
Joined: 23 Nov 08
Posts: 1073
Credit: 4,619,000,754
RAC: 353,615
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27598 - Posted: 6 Dec 2012 | 23:22:41 UTC - in response to Message 27591.

I have been crunching Nathans WUs with 2 GTX460 (1GB GDDR each if I remember well) on 2 different systems (Win and Ubuntu) and then on one system with both cards together : I never had this problem with Nathan's WUs.

Sorry but if you get problems , I' m pretty sure this is a problem with YOUR system.

upgrade or change your GPU or upgrade rour system ! !

Things have changed. The last time you connected with your 2 GTX 460 boxes was March/2011 and Feb/2012:

http://www.gpugrid.net/hosts_user.php?sort=rpc_time&rev=0&show_all=1&userid=57540

Now you're running a GTX 690 & a GTX 590. Your GTX 590 shows 28 "errors while computing", 27 of which are NATHAN WUs, and your current GPUs are among the ones that actually work better with these WUs. Upgrade our GPUs? They work very well with the other WUs. How about upgrade the programming...

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 668
Credit: 2,498,095,550
RAC: 12
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27599 - Posted: 6 Dec 2012 | 23:46:44 UTC - in response to Message 27591.


Sorry but if you get problems , I' m pretty sure this is a problem with YOUR system.

upgrade or change your GPU or upgrade rour system ! !


It's not a problem with my system and while you're no doubt sitting on piles of cash to upgrade everything so you can have your million plus RAC not everyone can or wants to spend their hard earned cash in these hard times.

Are you a Banker?
____________
Radio Caroline, the world's most famous offshore pirate radio station.
Great music since April 1964. Support Radio Caroline Team -
Radio Caroline

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27602 - Posted: 7 Dec 2012 | 7:05:48 UTC - in response to Message 27598.


Things have changed. The last time you connected with your 2 GTX 460 boxes was March/2011 and Feb/2012:

http://www.gpugrid.net/hosts_user.php?sort=rpc_time&rev=0&show_all=1&userid=57540

Now you're running a GTX 690 & a GTX 590. Your GTX 590 shows 28 "errors while computing", 27 of which are NATHAN WUs, and your current GPUs are among the ones that actually work better with these WUs. Upgrade our GPUs? They work very well with the other WUs. How about upgrade the programming...


The errors you are mentioning are real WUs errors , they did not even start !

It's not a problem with my system and while you're no doubt sitting on piles of cash to upgrade everything so you can have your million plus RAC not everyone can or wants to spend their hard earned cash in these hard times.

Are you a Banker?


How can you explain that, for the involved WUs, not everybody with this kind of GPU is affected. If a WU is badly programmed it will affect a lot of different kinds of GPU.

No sir, I'm not a banker ,lol , but some put their money in a house, other in a car etc... It' s just a matter of choice...
Upgrading to a High-end card is not a must and there are still a lot of gtx 5xx series cards above the GTX460 available as second hand.
Regards.



____________
Lubuntu 16.04.1 LTS x64

Profile Beyond
Avatar
Send message
Joined: 23 Nov 08
Posts: 1073
Credit: 4,619,000,754
RAC: 353,615
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27603 - Posted: 7 Dec 2012 | 13:00:20 UTC - in response to Message 27602.

How can you explain that, for the involved WUs, not everybody with this kind of GPU is affected.

Everyone with the GTX 460 786MB is affected with the 1/2 speed NATHAN bug. I've looked at the results from dozens of them. They're one of the more common GPUs being run and the NATHAN WU performance is severely degraded on all of them. Bad for the project, bad for the users

some put their money in a house, other in a car etc... It' s just a matter of choice...

So we're supposed to buy video cards instead of a house or car? I think your priorities are a bit out of whack. All that needs to be done is fix some WUs or set up a separate queue, either of which should be simple.

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27608 - Posted: 7 Dec 2012 | 18:57:40 UTC - in response to Message 27603.
Last modified: 7 Dec 2012 | 19:00:19 UTC

Everyone with the GTX 460 786MB is affected with the 1/2 speed NATHAN bug. I've looked at the results from dozens of them. They're one of the more common GPUs being run and the NATHAN WU performance is severely degraded on all of them. Bad for the project, bad for the users

Only all GTX460 with 768MB ?
What about those with 1GB ? Is somebody with 1GB GTX460 also affected ?
I don't think Nathan wants to consider reprogramming (if this is possible...) to fix the GTX460 with 768MB but if he does this is allright, but only if this is not a penalty for all others.
Nathan it's up to you , if you ever read this thread... (I hope so)

So we're supposed to buy video cards instead of a house or car? I think your priorities are a bit out of whack.

I did not say I'm living on the street and that I don't have a car. But I'm sure you have very well understood what I have been talking about..
Regards
____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27609 - Posted: 7 Dec 2012 | 20:09:59 UTC - in response to Message 27608.

Is somebody with 1GB GTX460 also affected ?


Oh.. YES!! I've been following this thread for MONTHS!! Since mid October, when I started recording all WUs, I've had 40 "GOOD" Nathans and 72 "BAD ONES".

If only Nathan would join us here and not sulk in the background... He MUST know about this thread!!!



____________

mikey
Send message
Joined: 2 Jan 09
Posts: 276
Credit: 352,977,990
RAC: 787,053
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 27620 - Posted: 8 Dec 2012 | 18:52:52 UTC - in response to Message 27608.

So we're supposed to buy video cards instead of a house or car? I think your priorities are a bit out of whack.

I did not say I'm living on the street and that I don't have a car. But I'm sure you have very well understood what I have been talking about..
Regards


I personally have ALMOST a 2 MILLION RAC at another project, AND live in a home and own an 8 year old car AND a 1 year old car, actually BOTH are suv's. I also have 15 pc's, all but 2 with gpu's crunching, running here at home and have a $500.00+ a month electric bill! My electric bill is over twice what my neighbor pays, he has a similar family and house size. It is called disposable income, some of us have it and some don't! IMHO I am not sure it matters though, we ALL crunch and THAT is what is important, not HOW MUCH we crunch.
____________

Bjarke
Send message
Joined: 1 Mar 09
Posts: 8
Credit: 74,871,366
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27655 - Posted: 11 Dec 2012 | 17:56:21 UTC

Regarding NATHAN WUs:
I believe the issue some of you are having is caused by high memory usage of Nathan WUs running on cards with less RAM. This is a hypothesis, but I have noticed the following:


  • TONI WUs (long) e.g. WU 3928763 causes 92% GPU load, and approx 500 MB GPU memory use.
  • NATHAN WUs (long) e.g. WU 6183375 causes 96-98% GPU load, and approx 1100 MB GPU memory use.


I am having no trouble at all running NATHAN WUs, but all of my cards have +2GB of memory. I can imagine that if your card has less than arround 1100 MB of RAM, you would face some problems like the ones you have with your 768MB GTX460. When I read the Recommended GPU FAQ I notice that the GTX460 is only recommended for use with short tasks, and that all GPU in the category "highly recommended" has 1GB of RAM or more.

Have you tried using GPU-Z utility to monitor the load and memory usage on your GPU?

TheFiend
Send message
Joined: 26 Aug 11
Posts: 98
Credit: 1,830,705,479
RAC: 961,552
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27663 - Posted: 12 Dec 2012 | 0:05:06 UTC - in response to Message 27655.

Regarding NATHAN WUs:
I believe the issue some of you are having is caused by high memory usage of Nathan WUs running on cards with less RAM. This is a hypothesis, but I have noticed the following:

  • TONI WUs (long) e.g. WU 3928763 causes 92% GPU load, and approx 500 MB GPU memory use.
  • NATHAN WUs (long) e.g. WU 6183375 causes 96-98% GPU load, and approx 1100 MB GPU memory use.


I am having no trouble at all running NATHAN WUs, but all of my cards have +2GB of memory. I can imagine that if your card has less than arround 1100 MB of RAM, you would face some problems like the ones you have with your 768MB GTX460. When I read the Recommended GPU FAQ I notice that the GTX460 is only recommended for use with short tasks, and that all GPU in the category "highly recommended" has 1GB of RAM or more.



I found the GTX460 768MB was fine with all long WU's except for NATHANS, which needed more memory to run.

Profile skgiven
Volunteer moderator
Project tester
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,925,443,810
RAC: 212,063
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27671 - Posted: 12 Dec 2012 | 14:09:01 UTC - in response to Message 27663.

Normally GPUGrid tasks scale down well; require less GDDR on smaller GPU's.
I don't have a GTX460, so I cannot check how much memory these tasks use.
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Profile nate
Send message
Joined: 6 Jun 11
Posts: 124
Credit: 2,928,865
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 27676 - Posted: 12 Dec 2012 | 17:04:30 UTC
Last modified: 12 Dec 2012 | 17:05:11 UTC

Sorry for the late reply. I did not know about this thread until a few days ago, and I just got around to replying now. The original subject matter wasn't directed at me, and so I wasn't following it.

As Bjarke and TheFiend have said, the likely issue is that the 460 does not have enough memory. My simulations are typically of larger biological systems than those of the other scientists, meaning more on-card memory is required for them. If the entire simulation can't fit on the card memory, a secondary algorithm must be used which is slower. There is not much we can do about this.

However, I don't think that it results in a 50% slowdown (1/2 speed). Tomba, for your case specifically, I have pulled some stats from the database (for the period from Nov6 to Nov19, the period I could query right now):

Total: 23 WUs returned, 12.7 hours average runtime, average 6605 credits/hour
Non-NATHAN tasks: 13 WUs returned, 11.7 hours average runtime, average 7186 credits/hour
NATHAN tasks: 10, 13.9 hours average runtime, average 5971 credits/hour

So, it looks like it causes a ~17% slowdown, not 50%. Obviously not insignificant, but there isn't anything I can do about it. A few percentage of that might be due to other factors as well. Since it only affect a small percentage of our users (those with low memory cards), and only causes partial slowdowns for them, we choose to proceed. The BOINC server doesn't allow a huge amount of leeway with regard to what machines get which WUs, so our only choice is to continue or not do the research at all.

Again, sorry for the delayed response.

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27677 - Posted: 12 Dec 2012 | 18:04:57 UTC - in response to Message 27676.

Hello Nathan! Welcome to our thread (at last!!!).

Sorry for the late reply. I did not know about this thread until a few days ago, and I just got around to replying now. The original subject matter wasn't directed at me, and so I wasn't following it.

There is a moderator posting to this thread. I would have thought that he/she would have notified the concerned scientist (you) in less than the five months this thread has been active.

As Bjarke and TheFiend have said, the likely issue is that the 460 does not have enough memory. My simulations are typically of larger biological systems than those of the other scientists, meaning more on-card memory is required for them. If the entire simulation can't fit on the card memory, a secondary algorithm must be used which is slower. There is not much we can do about this.

Slower... (?)

However, I don't think that it results in a 50% slowdown (1/2 speed).


That is not the issue here. I absolutely don't have a problem if my modest GPU hardware causes a slowdown of your WUs. What I and many other volunteers are experiencing is a lock-out of our PCs. Normal Windows operations take 10-times longer with some of your WUs.

Tomba,


That's me!

for your case specifically, I have pulled some stats from the database (for the period from Nov6 to Nov19, the period I could query right now):

Total: 23 WUs returned, 12.7 hours average runtime, average 6605 credits/hour
Non-NATHAN tasks: 13 WUs returned, 11.7 hours average runtime, average 7186 credits/hour
NATHAN tasks: 10, 13.9 hours average runtime, average 5971 credits/hour


That's all very well, but did you also notice that, in that time period, I aborted no less than 30 of your WUs because they were killing my PC performance?

Dear Nathan. Pretty please, have another look. Thank you, Tom
____________

TJ
Send message
Joined: 26 Jun 09
Posts: 815
Credit: 1,470,348,955
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27679 - Posted: 12 Dec 2012 | 18:36:56 UTC

I noticed that some NATHAN WU's result in stopping the CUDA program and thus error out. Mostly that are long runs.
However a lot of these NATHAN's run complete.
And I have also had TONI's that result in error by stopping (or killing) the CUDA program.
But I guess this happens once in a while as I see it also with other projects. It's all in the game, and this game is important.

I have a GTX285 and a GTX550Ti by the way.
____________
Greetings from TJ

Profile nate
Send message
Joined: 6 Jun 11
Posts: 124
Credit: 2,928,865
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 27685 - Posted: 13 Dec 2012 | 15:29:56 UTC - in response to Message 27677.
Last modified: 13 Dec 2012 | 15:31:24 UTC

I can't tell exactly what's going on with your machine since I'm not sitting there using it, but from what I can tell reading this thread the likely problems are:

(1) You have one GPU, which is naturally your display card as well. You are attempting to crunch GPU tasks and use the computer at the same time. This is generally not a good idea with an older card where your display requirements are heavy, especially when watching movies or doing 3D work, and (2) if there isn't enough memory on the card for the WU, then the program/OS will look to your system RAM, and then to the swap memory (Virtual memory on Windows). If it is doing the latter, performance will be catastrophically bad, because it is using the hard drive as RAM. It sounds like this might be what is happening if it is "killing" your machine.

To resolve the first problem, you should consider suspending GPU tasks while the computer is in use. To resolve the second, when the machine is not in use make sure that any programs that use significant memory, and don't need to be running, have been closed. This would hopefully keep large, efficient (98%+ gpu usage) work units such as mine from causing systems like yours problems. My final suggestion is for you to crunch only on the short queue. I do not have work units there, but even so I can't promise there are no large, memory heavy, efficient work units there either.

Please don't assume that I don't believe that you are having serious problems, or that it is not an issue on our side. However, if the problem goes beyond what is mentioned here, then it is truly a problem that can not be fixed anytime soon. For example, a memory leak caused by the application or the Nvidia driver would be an example. I have mentioned the possibility to the programmers here, but they are busy and an update would take time. We can only update the application once or twice a year for safety and continuity reasons. And obviously if it is an obscure driver error, we have no control over it.

If you have any additional information, such as what happens to the memory when you see this problem, then we might be able to help more, or at least resolve it in future versions of the application.

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27686 - Posted: 13 Dec 2012 | 16:35:47 UTC - in response to Message 27685.

Hello Nathan. Thank you for your insights.

(1) You have one GPU, which is naturally your display card as well. You are attempting to crunch GPU tasks and use the computer at the same time. This is generally not a good idea with an older card where your display requirements are heavy, especially when watching movies or doing 3D work

I don't do movies or 3D work.

(2) if there isn't enough memory on the card for the WU, then the program/OS will look to your system RAM, and then to the swap memory (Virtual memory on Windows).

I have one gig on the card, and six gigs of RAM driving an i7. Right now I'm running I6R9-NATHAN_RPS1120913_respawn2-65-100-RND8231. I don't have a performance problem. Here's what my gadgets tell me:



I guess I'll wait for a killer WU and report again.

FYI, in the past two months I have had one non-Nathan WU that killed my PC: 2CN4_3-GIANNI_bias-11-25-RND1144
____________

Profile skgiven
Volunteer moderator
Project tester
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,925,443,810
RAC: 212,063
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27688 - Posted: 13 Dec 2012 | 19:30:38 UTC - in response to Message 27679.

This thread started off as one of several threads about the 3.1 vs 4.2 app. It then went to SLI and while not quite hi-jacked, as the initial problem was solved (and the title fits), it’s moved onto what seems to be a general complaints thread, about performance issues and limitations of the GTX460, amongst other topics.

I think it’s important that the researchers specify the requirements for various task types, so people can opt in or out of the project, or for long or short tasks.
Of late we have had some heavy GPU memory requirements for several lines of research, some WU’s requiring too much CPU to facilitate even a mid-range GPU and untested release failures. Not much point adding Twitter if you can’t Tweet the basics in the forum!

I wasn't aware that there are presently work units requiring >800MB for normal system performance. These things are normally identified by testers, following testing. Mind you, nobody posted how much GPU memory their GPU’s were actually using, and you can’t see this in the standard error output.
Anyway, these are difficult situations to resolve without having some sort of task-selection mechanism.

If this is a bigger problem than GTX460’s only, I would suggest such heavy-memory or CPU overly-dependent tasks be run only from the Beta queue, or some other test queue for wanton/extreme tasks. While I commend expanding the boundaries of research, I would prefer more choice.

If the server excluded CC2.1 cards, then it would exclude all those GTX460's with 1GB GDDR, and other CC2.1 cards with large amounts of GPU memory. The situation is frustrating, and that wouldn't make it any less so. Perhaps CC2.1 for short tasks only?

tomba, W7 allocates 10% of the GPU’s memory to itself and allocates this ‘dynamically’. Historically, this Desktop Window Management system, results in an 11% or worse performance loss for GPUGrid tasks on W7/Vista, and that’s after disabling Aero and making system optimizations. Note 2008servers are around 4% and 2003 servers operate like an already optimized version of XP.

On my GTX470 a NATHAN_RPS* (non respawn) task uses 960MB GDDR5 on a 2008R2 server, as did the next such task. On W7 it required 1010MB. The GTX470 has 1280MB (roughly), so the 10% hit doesn't effect these WU's on a GTX470 on W7. However, 1GB less 10% would put your GTX460 1GB into the category of 'not enough GPU memory' (921MB), basically because of the Operating System. This would not be the case on XP or Linux and it's less on a server.
There are other considerations that could be at play, such as occasional downclocking, not having enough CPU free to support the GPU, having limited system memory, or running heavy disk I/O tasks, but unless you specify these details I can't make any such determination. The same applies for other moderators/testers and researchers.

From what you report, the respawn tasks use substantially less GPU memory, but there are at least two types of non-respawn NATHAN tasks. I don’t know if both, or just one has the memory requirement problem?

____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Profile nenym
Send message
Joined: 31 Mar 09
Posts: 137
Credit: 741,250,218
RAC: 591,741
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27690 - Posted: 13 Dec 2012 | 20:20:36 UTC - in response to Message 27688.

Perhaps CC2.1 for short tasks only?
I do NOT agree with you. I have got one of the first GTX560Ti CC2.1 1023MB and I have no problem on 64bit XP crunching any longrun task. Some teammates have GPU CC2.1 with memory over 1,2GB and they did not report any problem in our forum or here. By my point of view there is no reason to disable longrun tasks for CC2.1 due to problem of one or several specific systems.
In all I agree with you, especially in point of necessity to specify system having problems exactly.
Off Topic for tomba: How do you like WCG HCC GPU tasks on nVidia GPU?

Profile skgiven
Volunteer moderator
Project tester
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,925,443,810
RAC: 212,063
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27692 - Posted: 13 Dec 2012 | 22:08:43 UTC - in response to Message 27690.
Last modified: 13 Dec 2012 | 22:24:00 UTC

nenym, You are quite right, if the issue isn't just with 768MB cards (or less), then it could be with numerous 1GB cards (on Vista/W7) when running these particular NATHAN tasks.
While there probably isn't any problem with 1GB cards on Linux, XP, 2003 server, and possibly not on 2008R2 servers (touch and go with that one), it's almost certainly not isolated to GTX460 768 cards. It would certainly impact on the 512MB GT440 and GTS450 and probably most 1GB cards on W7/Vista including the GeForce GTX465 (1GB), four types of GTX460, two types of GTS450, some GT440's, GTX560Ti, GTX560, GTX560SE, GTX555, GTX550 Ti, GTX545, GT530, and most disturbingly it could cause problems with some GF600 cards; GT640, GTX650, and even the GTX650Ti.

Thanks,
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Profile nenym
Send message
Joined: 31 Mar 09
Posts: 137
Credit: 741,250,218
RAC: 591,741
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27693 - Posted: 13 Dec 2012 | 22:31:35 UTC - in response to Message 27692.

In that case I see one way you have mentioned yet - fourth queue for the video memory consumption longrun tasks. One could choose if run these tasks or not.

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27695 - Posted: 14 Dec 2012 | 7:53:53 UTC - in response to Message 27690.

Off Topic for tomba: How do you like WCG HCC GPU tasks on nVidia GPU?


Me no understand "WCG HCC"!!

____________

Profile nenym
Send message
Joined: 31 Mar 09
Posts: 137
Credit: 741,250,218
RAC: 591,741
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27701 - Posted: 14 Dec 2012 | 18:29:47 UTC - in response to Message 27695.

Off Topic for tomba: How do you like WCG HCC GPU tasks on nVidia GPU?


Me no understand "WCG HCC"!!
World Community Grid, project Help Conquer Cancer.

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 1961
Credit: 12,646,661,119
RAC: 7,299,965
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27702 - Posted: 14 Dec 2012 | 22:28:41 UTC - in response to Message 27695.

Off Topic for tomba: How do you like WCG HCC GPU tasks on nVidia GPU?


Me no understand "WCG HCC"!!

Then watch this.

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27703 - Posted: 14 Dec 2012 | 22:36:11 UTC - in response to Message 27701.

Hi everybody !

The title of this thread is "Very bad display performances" and was opened by me. It would be fine if I could add "with SLI and/or multi-GPU mode enabled"
For me this issue is SOLVED : If , when GPUGRID is running , your display refresh is VERY slow as if the system was slowed down, first uncheck SLI and/or multi-gpu performance modes if, of course, applicable to you.
If this is not sufficient or applicable, please open a NEW thread.

WU programming vs GPU capabilities is off topic.

Regards.

____________
Lubuntu 16.04.1 LTS x64

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27705 - Posted: 15 Dec 2012 | 13:25:27 UTC

I found what looks like a comprehensive GPU watcher; GPU Monitor:



Here we have two NATHANS with a TONI in the middle. None of then gave me PC lock-outs.

Left clock is CPU - little running in my i7 - and the right one shows less than 50% usage of my six gigs of RAM.

The left pair show almost identical GPU status. The other is using almost all my 1 gig of GPU RAM; only 13MB free. But it still did not interfere with normal PC processing.

I'm almost looking forward to killer NATHAN!!
____________

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27707 - Posted: 15 Dec 2012 | 18:18:56 UTC
Last modified: 15 Dec 2012 | 18:21:03 UTC

Oops - duplicate...
____________

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27715 - Posted: 17 Dec 2012 | 12:49:42 UTC - in response to Message 27705.

I'm almost looking forward to getting a killer NATHAN!!

I got one:



No effect on CPU or PC RAM, nor on most GPU measures. The only difference from the other three is that the GPU load is 99%. That seems to be what's killing my PC performance. It's nothing to do with GPU RAM (357MB free).
____________

Profile Beyond
Avatar
Send message
Joined: 23 Nov 08
Posts: 1073
Credit: 4,619,000,754
RAC: 353,615
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27724 - Posted: 17 Dec 2012 | 23:30:50 UTC - in response to Message 27676.

As Bjarke and TheFiend have said, the likely issue is that the 460 does not have enough memory. My simulations are typically of larger biological systems than those of the other scientists, meaning more on-card memory is required for them. If the entire simulation can't fit on the card memory, a secondary algorithm must be used which is slower. There is not much we can do about this.

However, I don't think that it results in a 50% slowdown (1/2 speed).

In my tests it's almost exactly a 50% slowdown on the GTX 460/786MB cards. It's a memory size problem since the GTX 460/1024MB cards do not have the 50% performance hit. ON the longer NATHAN WUs there's no way to make the 24hr deadline with cards with less than 1024MB either. The TONI WUs run in 12.5-13.5 hours on my 4 GTX 460/786MB cards. As suggested, a solution would be a separate queue for WUs requiring a large memory footprint. Should be simple to implement.

TJ
Send message
Joined: 26 Jun 09
Posts: 815
Credit: 1,470,348,955
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27727 - Posted: 18 Dec 2012 | 11:41:04 UTC

Well I have a GTX285 with 1Gb of memory, Vista x64 ultimate and driver 285.62.
This pc does short runs and long runs from Toni and Nathan and finish them between 65.000 and 78.000 seconds. Perhaps a slow down but finish with good results.

I had updated to latest driver and that resulted in a lot of errors so I downgraded again.
Now I have sometimes a problem that BOINC seems to hang and no respons of mouse of keyboard is possible.
When suspending GPUGRID, same happened and was Rosetta crunching. Booting and Rosetta suspending, GPUGRID running, also freeze after a few minutes.

Booted again, cleared my register, did a full virus scan and removed a piece of software (3.3 was the name) what I didn't know what it was or did. But afterwards the systems in crunching Rosetta and GPUGRID for already 23 hours and 8 minutes without issues or freezes.

For Tomba: perhaps you can check with process explorer (from Microsoft, free to download) if other applications are occupying your system.
____________
Greetings from TJ

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27728 - Posted: 18 Dec 2012 | 15:46:40 UTC - in response to Message 27727.

For Tomba: perhaps you can check with process explorer (from Microsoft, free to download) if other applications are occupying your system.

I think that the four examples I've posted of my system state when GPUGRID is running amply demonstrate that nothing unknown is occupying my system; pratically no CPU activity and <50% system RAM usage looks good to me!

But I will give Process Explorer a whirl...

Tomba
____________

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27729 - Posted: 18 Dec 2012 | 16:05:07 UTC - in response to Message 27728.

But I will give Process Explorer a whirl...

Everything that's running either I started or I know why it's running.

____________

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27730 - Posted: 18 Dec 2012 | 17:10:38 UTC - in response to Message 27724.

As suggested, a solution would be a separate queue for WUs requiring a large memory footprint. Should be simple to implement.

YES - YES - YES!! Let's do it, GPUGRID team! Pretty please...

____________

Profile nate
Send message
Joined: 6 Jun 11
Posts: 124
Credit: 2,928,865
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 27734 - Posted: 19 Dec 2012 | 16:53:36 UTC

Ok, I think a lot has come out of this thread, both with the original troubleshooting about display performances and the user concerns about performance and hardware capabilities. I am heading off on holidays tomorrow, so here are my final comments, at least for this thread.

Tomba and Jihal (et al.), thanks for the exhaustive work on trying to figure out the display issues and machine problems. Tomba, those posts with the full GPU stats are very helpful. I am not sure what it means yet, but it helps narrow down causes.

(skgiven) I think it’s important that the researchers specify the requirements for various task types, so people can opt in or out of the project, or for long or short tasks. Of late we have had some heavy GPU memory requirements for several lines of research, some WU’s requiring too much CPU to facilitate even a mid-range GPU and untested release failures. Not much point adding Twitter if you can’t Tweet the basics in the forum!

(Beyond) As suggested, a solution would be a separate queue for WUs requiring a large memory footprint. Should be simple to implement.


I think these are all valid concerns, and we'll discuss how to address them in the new year. Making modifications to how we send tasks or adding a new queue (or however we address the issue), is not a trivial task, since it requires we modify the server. Modifying the server can cause it to crash, or it can lead to loss of data. "Small" changes can turn into catastrophic failures, and we'll have to test them first.

The high memory usage NATHAN work units will disappear from the queue in the next week or two, so that will no longer be a problem. I will submit some more WUs soon, but they will not have high memory or cpu usage. We'll try to figure out how to handle those types of work units better in the future.

Now I'm off for the holidays. Merry Christmas, happy holidays, and happy new year to everyone!

tomba
Send message
Joined: 21 Feb 09
Posts: 497
Credit: 700,690,702
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27735 - Posted: 19 Dec 2012 | 17:22:31 UTC - in response to Message 27734.

Now I'm off for the holidays. Merry Christmas, happy holidays, and happy new year to everyone!

Enjoy, Nathan! Thank you for being on our side. Looking forward to another successful year for GPUGRID!

Tom

____________

Post to thread

Message boards : Number crunching : Very bad display performances