Advanced search

Message boards : Graphics cards (GPUs) : Nvidia 190.38, Cuda 2.3

Author Message
Mark Henderson
Send message
Joined: 21 Dec 08
Posts: 51
Credit: 26,320,167
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 11168 - Posted: 18 Jul 2009 | 8:09:58 UTC
Last modified: 18 Jul 2009 | 8:11:44 UTC

Just installed 190.38 Beta on Win. XP64 with 2 260s. Running good here. This has Cuda 2.3 also.

MarkJ
Volunteer moderator
Volunteer tester
Send message
Joined: 24 Dec 08
Posts: 738
Credit: 200,909,904
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11169 - Posted: 18 Jul 2009 | 9:06:42 UTC - in response to Message 11168.

Just installed 190.38 Beta on Win. XP64 with 2 260s. Running good here. This has Cuda 2.3 also.


Ahh so they did finally get a cuda 2.3 out (even if its still in beta). I wonder if the project admins will just skip to cuda 2.3.
____________
BOINC blog

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11170 - Posted: 18 Jul 2009 | 9:35:35 UTC - in response to Message 11169.

I wouldn't have thought so given they haven't rolled out 2.2 yet because of problems with some machines.
____________
Radio Caroline, the world's most famous offshore pirate radio station.
Great music since April 1964. Support Radio Caroline Team -
Radio Caroline

MarkJ
Volunteer moderator
Volunteer tester
Send message
Joined: 24 Dec 08
Posts: 738
Credit: 200,909,904
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11171 - Posted: 18 Jul 2009 | 9:37:50 UTC

Here are some notes about the cuda 2.3 beta that I managed to find on the nvidia web site

Jun 20 2009, 06:17 AM

The CUDA Toolkit and SDK 2.3 betas are now available to registered developers. They include the following features:

The CUFFT Library now supports double-precision transforms and includes significant performance improvements for single-precision transforms as well. See the CUDA Toolkit release notes for details.

The CUDA-GDB hardware debugger is now available for all supported Linux platforms and is included in the CUDA Toolkit installer.

GPUs in an SLI group are now enumerated individually, so you can achieve multi-GPU performance even when SLI is enabled for graphics.

New support for fp16 <-> conversion intrinsics allows storage of data in fp16 format with computation in fp32. Use of fp16 format is ideal for applications that require higher numerical range than 16-bit integer but less precision than fp32 and reduces memory space and bandwidth consumption.

The CUDA SDK has been updated to include:
A new pitchLinearTexure code sample that shows how to effeciently texture from pitch linear memory.

A new PTXJIT code sample illustrating how to use cuModuleLoadDataEx() to load PTX source from memory instead of loading a file.

Two new code samples for Windows, showing how to use the NVCUVID library to decode MPEG-2, VC-1, and H.264 content and pass frames to OpenGL or Direct3D for display.

Updated code samples showing how to properly align CUDA kernel function parameters so the same code works on both x32 and x64 systems.

The Visual Profiler (packaged separately) includes several enhancements:
All memory transfer API calls are now reported

Support for profiling multiple contexts per GPU.

Synchronized clocks for requested start time on the CPU and start/end times on the GPU for all kernel launches and memory transfers.

Global memory load and store efficiency metrics for GPUs with compute capability 1.2 and higher.

The CUDA Driver for MacOS is now packaged separately from the CUDA Toolkit.
Support for major Linux distros, MacOS X, and Windows:

Fedora 10, RHEL 4.7 & 5.3, SLED 10.2 & 11.0, OpenSUSE 11.1, and Ubuntu 8.10 & 9.04
Windows XP/Vista/7 with Visual Studio 8 (VC2005) and 9 (VC2008)
MacOS X 10.5.6 and later (32-bit)



____________
BOINC blog

Profile K1atOdessa
Send message
Joined: 25 Feb 08
Posts: 249
Credit: 370,320,941
RAC: 0
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11173 - Posted: 18 Jul 2009 | 11:28:13 UTC - in response to Message 11171.

GPUs in an SLI group are now enumerated individually, so you can achieve multi-GPU performance even when SLI is enabled for graphics.


That's a nice feature, given all the confusion from new members that have SLI enabled and can't figure out why there aren't multiple WU's running. Kind of a "use SLI when playing a game, but not when crunching without having to do anything" sort of thing.

Anyone tested this new feature yet?

Profile [BOINC@Poland]AiDec
Send message
Joined: 2 Sep 08
Posts: 53
Credit: 9,213,937
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwat
Message 11213 - Posted: 20 Jul 2009 | 22:32:57 UTC - in response to Message 11173.
Last modified: 20 Jul 2009 | 22:34:45 UTC

I`ve tested new drivers and I`m very satisfied - fast and fully stable (tested with other CUDA-software too). I wasn`t so satisfied since 178.24.

So at this moment I`d like to recommend 190.38 beta as best drivers for now.



It`s ofc my opinion and I just want to share it to the others. GL guys.
____________

popandbob
Send message
Joined: 18 Jul 07
Posts: 67
Credit: 40,277,822
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11240 - Posted: 22 Jul 2009 | 6:28:58 UTC - in response to Message 11213.

I`d like to recommend 190.38 beta


Just released as non beta :)

Bob

zpm
Avatar
Send message
Joined: 2 Mar 09
Posts: 159
Credit: 13,639,818
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11245 - Posted: 22 Jul 2009 | 12:01:51 UTC - in response to Message 11240.
Last modified: 22 Jul 2009 | 12:07:17 UTC

already found a bug, my 32-bit windows xp SP3 "9400 GT PCI" pc was doing something funky with pogo games, i rolled back the driver(June) and it's now corrected, but i don't have the power to report the bug to nvidia... can one of the reg. developer please get word to them... it wasn't java... it was the driver.

Profile Jet
Send message
Joined: 14 Jun 09
Posts: 25
Credit: 5,835,455
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 11309 - Posted: 25 Jul 2009 | 18:07:33 UTC - in response to Message 11168.

Hi, one question: did you notice some changes in perfomance ?
Unfortunately, installation of new 190.38 CUDA 2,3 was unlucky for me - new version doesn't recognize my second GTX260, so, I was roll them back to the previous.

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11312 - Posted: 25 Jul 2009 | 20:57:31 UTC
Last modified: 25 Jul 2009 | 20:59:08 UTC

Are you guys actually installing the CUDA 2.3 or Previous version of CUDA, I've never done that. All I've ever done is install whatever NVIDIA Drivers I needed to and ran the WU's. I've always thought the actual CUDA Program & Tools was just for Developers and not needed to run the WU's which I'm doing ...

I did install the new 190.38 Drivers on 7 Win XP Pro 64-Bit Box's this morning with no problems & running fine so far ... :)

zpm
Avatar
Send message
Joined: 2 Mar 09
Posts: 159
Credit: 13,639,818
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11313 - Posted: 25 Jul 2009 | 21:20:53 UTC - in response to Message 11312.

yea we install the toolkit...

popandbob
Send message
Joined: 18 Jul 07
Posts: 67
Credit: 40,277,822
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11318 - Posted: 26 Jul 2009 | 0:19:04 UTC

Jet, Did you install the 190.38 drivers on the second card as well?
The best way to make sure the drivers are installed on both cards is to update drivers via device manager.

zpm, No need to install the toolkit unless your developing a GPU app.

Bob

Profile Jet
Send message
Joined: 14 Jun 09
Posts: 25
Credit: 5,835,455
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 11320 - Posted: 26 Jul 2009 | 2:56:06 UTC - in response to Message 11318.

Thanks for afvice, drivers was installed for both...

naja002
Avatar
Send message
Joined: 25 Sep 08
Posts: 111
Credit: 10,352,599
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwat
Message 11321 - Posted: 26 Jul 2009 | 4:57:42 UTC

I had nothing but issues with 190.38 on an i7 920 running 2x 8800GSs and Vista 64 SP1---until I upgraded a couple of days ago from 6.6.36 to 6.6.37....then the issues went away. Yesterday I moved back to RT, forced the driver and 3D performance and everything seems fine as far as Gpugrid goes. The main issue I have now seems to have appeared when I moved up to firefox 3.5.1, but it's not affecting the grid.

Put 190.38 on a Q6600 2x gpu rig running Vista 64 SP1 and haven't had any problems. I just checked and it's still running 6.6.28! lol

Profile Paul D. Buck
Send message
Joined: 9 Jun 08
Posts: 1050
Credit: 37,321,185
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11323 - Posted: 26 Jul 2009 | 7:05:03 UTC

Ok, so now the question is this: Is it better on GPU Grid to use SLI or not?

Has anyone tested this yet?

I just installed 190.38 on one system and started it up so I will see if t works at all ... (3 minutes in and so far so good). Though I am not sure if we are running the newer application or not.

MarkJ
Volunteer moderator
Volunteer tester
Send message
Joined: 24 Dec 08
Posts: 738
Credit: 200,909,904
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11328 - Posted: 26 Jul 2009 | 9:13:07 UTC - in response to Message 11323.

Ok, so now the question is this: Is it better on GPU Grid to use SLI or not?

Has anyone tested this yet?

I just installed 190.38 on one system and started it up so I will see if t works at all ... (3 minutes in and so far so good). Though I am not sure if we are running the newer application or not.


At this point the app is still delivering the cuda 2.1 DLL's. As per GDF's message thread it should get to cuda 2.2 by the end of this week.

My understanding is the SLI context is part of cuda 2.3, but I only got this from reading the release notes.
____________
BOINC blog

frankhagen
Send message
Joined: 18 Sep 08
Posts: 65
Credit: 3,037,414
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 11333 - Posted: 26 Jul 2009 | 11:54:13 UTC

after upggrading yesterday evening i had the first big fat blue one after months of crunching without issues.. :(

as i can hear from my team-mates, there are several boxes which went down..

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11334 - Posted: 26 Jul 2009 | 12:43:57 UTC

I have 2 Box's out of 7 that have returned nothing but error's since Upgrading to the 190.38's, the other 5 Box's only show 1 error total which could have been from something else.

I checked the 2 Box's in question and didn't see anything wrong with them but backed off the Shaders 25 Points & will see if they run better. Neither one of the Box's was returning errors until the Driver Upgrade so I would suspect something is amiss with the Driver on some Box's.

I checked the Task Manger & it shows the 190.38 Driver installed on Both Cards in both Box's but if they keep getting the error will do a complete uninstall & reinstall of the Drivers on both cards again.

MarkJ
Volunteer moderator
Volunteer tester
Send message
Joined: 24 Dec 08
Posts: 738
Credit: 200,909,904
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11335 - Posted: 26 Jul 2009 | 13:34:31 UTC - in response to Message 11334.

I have 2 Box's out of 7 that have returned nothing but error's since Upgrading to the 190.38's, the other 5 Box's only show 1 error total which could have been from something else.

I checked the 2 Box's in question and didn't see anything wrong with them but backed off the Shaders 25 Points & will see if they run better. Neither one of the Box's was returning errors until the Driver Upgrade so I would suspect something is amiss with the Driver on some Box's.

I checked the Task Manger & it shows the 190.38 Driver installed on Both Cards in both Box's but if they keep getting the error will do a complete uninstall & reinstall of the Drivers on both cards again.


I started another thread here for errors.

I too have one machine failing on all its wu. It has GTX260's, so maybe its limited to the G200 chipset as I have 4 others with GTS250's (G82 chipset) working fine.
____________
BOINC blog

Mark Henderson
Send message
Joined: 21 Dec 08
Posts: 51
Credit: 26,320,167
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 11340 - Posted: 26 Jul 2009 | 14:27:50 UTC
Last modified: 26 Jul 2009 | 14:35:49 UTC

Am I correct in that Cuda 2.3 has to have the 190.. drivers but Cuda 2.2 "as we are going to" works with the 185..186.. series drivers? I dont believe 190 drivers are necessary for Cuda 2.2 only for Cuda 2.3..
According to Nvidia release highlights.

I know the Mod. said 190.. was necessary but I wonder if he was mistaken.

Profile Kokomiko
Avatar
Send message
Joined: 18 Jul 08
Posts: 190
Credit: 24,093,690
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11345 - Posted: 26 Jul 2009 | 17:31:14 UTC

After installing the driver 190.38 I have some heavy screen lags while playing games on my GTX295 with and without SLI mode. I've not seen this behavior in the last months.
____________

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11348 - Posted: 26 Jul 2009 | 19:20:30 UTC - in response to Message 11340.
Last modified: 26 Jul 2009 | 19:22:26 UTC

Am I correct in that Cuda 2.3 has to have the 190.. drivers but Cuda 2.2 "as we are going to" works with the 185..186.. series drivers? I dont believe 190 drivers are necessary for Cuda 2.2 only for Cuda 2.3..
According to Nvidia release highlights.

I know the Mod. said 190.. was necessary but I wonder if he was mistaken.


If it was a Mistake it's cost me over 24 hours already on 4 200 Series NVIDIA Cards because all they've done since installing the 190.38 Drivers is Toast the Wu's after about 1/2 hour of running ... I've had both Dual GPU Core Box's shut down for the last 4 hr's cause didn't know what else to do with them and was tired of seeing all the errors come off them.

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11358 - Posted: 27 Jul 2009 | 7:29:15 UTC - in response to Message 11340.
Last modified: 27 Jul 2009 | 8:12:29 UTC

Am I correct in that Cuda 2.3 has to have the 190.. drivers but Cuda 2.2 "as we are going to" works with the 185..186.. series drivers? I dont believe 190 drivers are necessary for Cuda 2.2 only for Cuda 2.3..
According to Nvidia release highlights.



You're correct.
____________
Radio Caroline, the world's most famous offshore pirate radio station.
Great music since April 1964. Support Radio Caroline Team -
Radio Caroline

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 11360 - Posted: 27 Jul 2009 | 8:10:11 UTC - in response to Message 11358.

Sure, you can install the CUDA2.2 drivers as well. They will work, that's the reason to release the new application as cuda2.2.
On the other hand, we have tested here 190.xx and they seem to work.

gdf

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11365 - Posted: 27 Jul 2009 | 9:44:45 UTC - in response to Message 11360.

Sure, you can install the CUDA2.2 drivers as well. They will work, that's the reason to release the new application as cuda2.2.
On the other hand, we have tested here 190.xx and they seem to work.

gdf


So far they've seemed to have worked on 4 Box's of mine with around 10 GPU Cores in them but on 3 Box's they haven't with 7 GPU Cores. 2 of those Box's with 4 Cores have turned in nothing but errors for about the last 40 hr's now. Not 1 good Wu has been returned off them since going to the 190.xx Drivers. The other Box has 3 GPU Cores in it & 1 of the cores or 2 will turn in an error every now & then since updating to the 190.38 Drivers.

I've tried Marks RivaTuner Fix for the Cards dropping their Speed but that hasn't worked. They keep resetting themselves to lower speeds within an hour or so after getting them back up to speed again. I reinstalled the 186.18 Drivers too on both those Boxes but they continue to get the errors now where they weren't getting any before I updated them to the 190.38 Drivers.

When you refer to the CUDA2.2 Drivers & the 190.xx Drivers are you refering to different Drivers or the same thing. All I have installed is the NVIDIA Drivers but not CUDA2.2 Drivers. Do you think it would help to install the CUDA2.2 Drivers if they are different from the 190.xx Drivers ???


Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11366 - Posted: 27 Jul 2009 | 9:53:17 UTC - in response to Message 11365.

[When you refer to the CUDA2.2 Drivers & the 190.xx Drivers are you refering to different Drivers or the same thing. All I have installed is the NVIDIA Drivers but not CUDA2.2 Drivers. Do you think it would help to install the CUDA2.2 Drivers if they are different from the 190.xx Drivers ???


Nvidia drivers gives your card the capability to run 2.2 apps but GPUGRID has to compile their apps with 2.2 to enable you to run GPUGRID using 2.2 runtime.
Same goes for 190.xxx drivers they give you capability to run 2.3 runtime but only on apps that have been compiled to do so.
____________
Radio Caroline, the world's most famous offshore pirate radio station.
Great music since April 1964. Support Radio Caroline Team -
Radio Caroline

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11368 - Posted: 27 Jul 2009 | 10:10:52 UTC - in response to Message 11366.
Last modified: 27 Jul 2009 | 10:13:00 UTC

Has everyone tried upgrading to BOINC 6.6.37 so that we're all singing from the same hymn sheet?

This has a significant fix for CUDA apps especially the use of video ram. I am running it no problems for days now.

XP32

XP64
____________
Radio Caroline, the world's most famous offshore pirate radio station.
Great music since April 1964. Support Radio Caroline Team -
Radio Caroline

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11369 - Posted: 27 Jul 2009 | 10:18:26 UTC - in response to Message 11368.

Has everyone tried upgrading to BOINC 6.6.37 so that we're all singing from the same hymn sheet?

This has a significant fix for CUDA apps especially the use of video ram. I am running it no problems for days now.

XP32

XP64


I'll give the 6.6.37 a try on those Box's in awhile, right now their useless to me to run the Grid Project so anythings worth a try.

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11372 - Posted: 27 Jul 2009 | 10:51:24 UTC - in response to Message 11369.

I'll give the 6.6.37 a try on those Box's in awhile, right now their useless to me to run the Grid Project so anythings worth a try.


I would also recommend that anyone having trouble with the 190.xxx drivers uninstalls all Nvidia drivers including PhysX REBOOT and then installing the new drivers as a clean install ensuring you have anti virus turned off. I appologise to anyone who has already done this.
____________
Radio Caroline, the world's most famous offshore pirate radio station.
Great music since April 1964. Support Radio Caroline Team -
Radio Caroline

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11374 - Posted: 27 Jul 2009 | 14:36:25 UTC - in response to Message 11372.
Last modified: 27 Jul 2009 | 14:40:47 UTC

I'll give the 6.6.37 a try on those Box's in awhile, right now their useless to me to run the Grid Project so anythings worth a try.


I would also recommend that anyone having trouble with the 190.xxx drivers uninstalls all Nvidia drivers including PhysX REBOOT and then installing the new drivers as a clean install ensuring you have anti virus turned off. I appologise to anyone who has already done this.


Did all that & installed the 6.6.37 Client along with the 190.38 Drivers and have 2 Wu Errors already on 2 different Box's, the same 2 Box's that have been giving 100% errors for the last 48 hours now. This all started with the original installation of the 190.38 Drivers, up until that time none of the Cards was giving errors but now that they've started there's been no way to stop them so far

Not ever reverting back to the 185.18 Drivers has helped, so I'm out 6 GPU Cards because of a Driver Change, just great. Now the one Box won't even get anymore Wu's because it's used up it's Daily Quota so I can't try anything else until tonight when the Quota's reset ...

Profile Hydropower
Avatar
Send message
Joined: 3 Apr 09
Posts: 70
Credit: 6,003,024
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwat
Message 11375 - Posted: 27 Jul 2009 | 14:45:20 UTC - in response to Message 11374.

Hi, is that the dual-GTX295 box that has the 01 error ? I noticed that you still have them overclocked (Clock rate: 1476000 kilohertz). What I find strange with the nvidia drivers and the 'precision' tool from evga is that you can test a speed and it will say "*DRIVER* has passed speed check". This makes me suspicious of whether the *driver* is a serious bottleneck for overclocking. I would expect a message like "New speed accepted by *CARD*) or similar to make it clear that overclocking does not depend on a driver revision.

So... You might want to try the stock clocks and see how that works and clock up from there. Good luck.
____________
Join team Bletchley Park, the innovators.

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 11376 - Posted: 27 Jul 2009 | 15:15:43 UTC - in response to Message 11374.

Which hostid is out of quota? I will clear it.

gdf

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11377 - Posted: 27 Jul 2009 | 15:25:44 UTC

ID: 17909 ... Thanks GDF, I'm going to set the clock back to default & see what happens, if that doesn't work I'll try each card seperatly. Usually it's just one of the Cards hats causing both of them to get errors or so thats what I've found in the past ...

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11379 - Posted: 27 Jul 2009 | 15:28:34 UTC - in response to Message 11378.

Hi, is that the dual-GTX295 box that has the 01 error ? I noticed that you still have them overclocked (Clock rate: 1476000 kilohertz). What I find strange with the nvidia drivers and the 'precision' tool from evga is that you can test a speed and it will say "*DRIVER* has passed speed check". This makes me suspicious of whether the *driver* is a serious bottleneck for overclocking. I would expect a message like "New speed accepted by *CARD*) or similar to make it clear that overclocking does not depend on a driver revision.

So... You might want to try the stock clocks and see how that works and clock up from there. Good luck.



Yes, I didn't know it was getting that many errors though until I just looked at it. It must be going down hill fast as it wasn't getting that many before. I'll go drop that back to stack speeds too & see what happens.

popandbob
Send message
Joined: 18 Jul 07
Posts: 67
Credit: 40,277,822
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11412 - Posted: 28 Jul 2009 | 2:23:49 UTC - in response to Message 11375.

'precision' tool from evga is that you can test a speed and it will say "*DRIVER* has passed speed check"


This test only finds out if the GPU can handle the Overclock.
It doesn't test for the actual stability.
Bob

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11481 - Posted: 29 Jul 2009 | 18:58:49 UTC - in response to Message 11374.

Not even reverting back to the 186.18 Drivers has helped, so I'm out 6 GPU Cards because of a Driver Change, just great.


We had cases before where people upgraded the driver to some version, got only errors and upon reverting to the previously working drivers they still got only errors.

MrS
____________
Scanning for our furry friends since Jan 2002

Profile Hydropower
Avatar
Send message
Joined: 3 Apr 09
Posts: 70
Credit: 6,003,024
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwat
Message 11485 - Posted: 29 Jul 2009 | 21:12:53 UTC - in response to Message 11481.

@MrS: I would then presume that in those cases there were remains of the new drivers in there. I nowadays make ghost images of my system before I change anything major. It just takes too much time and hassle to revert in case of issues otherwise. Now at least I can play back the ghost onto the drive and go.

@popandbob: I realize it doesn't test actual stability, I was more putting the focus on the word DRIVER. It seems illogical to me to refer to testing a driver if in fact you are testing the GPU, unless one actually means to indicate that the driver was tested and not (only) the GPU. I can imagine that a driver has influence on the way the GPU can handle high command speeds (from my haydays of direct I/O assembly programming on ATI cards).
____________
Join team Bletchley Park, the innovators.

Mark Henderson
Send message
Joined: 21 Dec 08
Posts: 51
Credit: 26,320,167
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 11499 - Posted: 30 Jul 2009 | 11:29:19 UTC
Last modified: 30 Jul 2009 | 11:30:29 UTC

Hello Admin,
Would it be possible to do as Seti and have it to where the 2 DLL files for Cuda 2.3 could be " dropped " into the GPU folder and replace the 2.2 files?
This is possible on the Seti optimized apps. and works great.
For the people that 190.. drivers are running problem free this could give a small added boost.
I tried it one time and it would not work because the "file size " was incorrect and it errored out.
If the software did not look at the file size it would maybe work.
It may also help the developers get a jump on any issues with 2.3 also.

Just a question.
Thanks

Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 1576
Credit: 5,600,136,851
RAC: 8,787,368
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11504 - Posted: 30 Jul 2009 | 16:11:10 UTC - in response to Message 11499.

Hello Admin,
Would it be possible to do as Seti and have it to where the 2 DLL files for Cuda 2.3 could be " dropped " into the GPU folder and replace the 2.2 files?
This is possible on the Seti optimized apps. and works great.
For the people that 190.. drivers are running problem free this could give a small added boost.
I tried it one time and it would not work because the "file size " was incorrect and it errored out.
If the software did not look at the file size it would maybe work.
It may also help the developers get a jump on any issues with 2.3 also.

Just a question.
Thanks

That sounded like an interesting question, so I'm giving it a try on two hosts.

Preliminary observation - makes no difference at all, but both tasks are still below 10% - it'll be tomorrow morning (UTC) before I get a definitive answer. I'm running this on my two Q6600/9800GT hosts (XP SP3), if anyone wants to watch.

I can only presume that GPUGrid relies much less on Fourier Transforms than SETI - it was fft that made the difference at SETI.

OldChap
Send message
Joined: 5 Jul 09
Posts: 14
Credit: 28,617,272
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwat
Message 11514 - Posted: 30 Jul 2009 | 19:02:26 UTC

Gentlemen, I am new to gpugrid having folded for some months...May I ask...when you are talking of these errors since upgrading drivers ...are you getting the driver failed and recovered message??

zpm
Avatar
Send message
Joined: 2 Mar 09
Posts: 159
Credit: 13,639,818
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11519 - Posted: 30 Jul 2009 | 19:54:51 UTC - in response to Message 11514.
Last modified: 30 Jul 2009 | 19:56:28 UTC

some of us do but that has be deduced to problems with the driver itself and posted on nvidia...

were talking about app errors wit the app and boinc...

and some these are related to the fact that gpugrid is still built upon 2.1 cuda which is in the upgrade department right now.

OldChap
Send message
Joined: 5 Jul 09
Posts: 14
Credit: 28,617,272
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwat
Message 11525 - Posted: 30 Jul 2009 | 22:18:49 UTC - in response to Message 11519.

Thanks for that info..would you then be interested in the fact that rolling back to 182.08 then 185.85,186's and 190's they all display the same behaviour
Forcing Riva and locking 3D has no impact nor does 6.6.36 to 6.6.37.(this on vista 64 and 2* gtx260 (192's) Cuda 2.1.

Perhaps the most annoying thing for me though is that errors can occur from just a few seconds out to almost 100% complete.

I think I may just try with xp again.

Mike

Mark Henderson
Send message
Joined: 21 Dec 08
Posts: 51
Credit: 26,320,167
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 11527 - Posted: 31 Jul 2009 | 0:00:52 UTC - in response to Message 11504.

It worked on mine until it downloaded a new task and it errored out on the download " Cuda.dll Unexpected size " or something like that.

Hello Admin,
Would it be possible to do as Seti and have it to where the 2 DLL files for Cuda 2.3 could be " dropped " into the GPU folder and replace the 2.2 files?
This is possible on the Seti optimized apps. and works great.
For the people that 190.. drivers are running problem free this could give a small added boost.
I tried it one time and it would not work because the "file size " was incorrect and it errored out.
If the software did not look at the file size it would maybe work.
It may also help the developers get a jump on any issues with 2.3 also.

Just a question.
Thanks

That sounded like an interesting question, so I'm giving it a try on two hosts.

Preliminary observation - makes no difference at all, but both tasks are still below 10% - it'll be tomorrow morning (UTC) before I get a definitive answer. I'm running this on my two Q6600/9800GT hosts (XP SP3), if anyone wants to watch.

I can only presume that GPUGrid relies much less on Fourier Transforms than SETI - it was fft that made the difference at SETI.

Profile koschi
Avatar
Send message
Joined: 14 Aug 08
Posts: 124
Credit: 466,579,198
RAC: 362,756
Level
Gln
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 11529 - Posted: 31 Jul 2009 | 5:18:27 UTC - in response to Message 11527.

http://boincfaq.mundayweb.com/index.php?language=1&view=91

Please check if you can convince BOINC with this option to ignore the incorrect size and proceed with 2.3 libs:

<dont_check_file_sizes>0|1</dont_check_file_sizes>

Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 1576
Credit: 5,600,136,851
RAC: 8,787,368
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11531 - Posted: 31 Jul 2009 | 7:25:38 UTC

First task with 2.3 runtime DLLs (1041263) errored out after 14 hours with "Cuda error: Kernel [frc_sum_kernel_dihed] failed in file 'force.cu' in line 313 : unspecified launch failure."

Task on second machine, and replacement for this one, still running: but proceed with caution.

MarkJ
Volunteer moderator
Volunteer tester
Send message
Joined: 24 Dec 08
Posts: 738
Credit: 200,909,904
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11540 - Posted: 31 Jul 2009 | 11:50:06 UTC - in response to Message 11527.

It worked on mine until it downloaded a new task and it errored out on the download " Cuda.dll Unexpected size " or something like that.

Hello Admin,
Would it be possible to do as Seti and have it to where the 2 DLL files for Cuda 2.3 could be " dropped " into the GPU folder and replace the 2.2 files?
This is possible on the Seti optimized apps. and works great.
For the people that 190.. drivers are running problem free this could give a small added boost.
I tried it one time and it would not work because the "file size " was incorrect and it errored out.
If the software did not look at the file size it would maybe work.
It may also help the developers get a jump on any issues with 2.3 also.

Just a question.
Thanks

That sounded like an interesting question, so I'm giving it a try on two hosts.

Preliminary observation - makes no difference at all, but both tasks are still below 10% - it'll be tomorrow morning (UTC) before I get a definitive answer. I'm running this on my two Q6600/9800GT hosts (XP SP3), if anyone wants to watch.

I can only presume that GPUGrid relies much less on Fourier Transforms than SETI - it was fft that made the difference at SETI.



You'd need to use an app_info file. That way you'd bypass BOINC checking the signatures on the files.
____________
BOINC blog

Mark Henderson
Send message
Joined: 21 Dec 08
Posts: 51
Credit: 26,320,167
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 11543 - Posted: 31 Jul 2009 | 12:22:16 UTC

Would just adding this line into the CC_Config file work?

<dont_check_file_sizes>0|1</dont_check_file_sizes>

Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 1576
Credit: 5,600,136,851
RAC: 8,787,368
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11544 - Posted: 31 Jul 2009 | 13:08:06 UTC - in response to Message 11540.

I can only presume that GPUGrid relies much less on Fourier Transforms than SETI - it was fft that made the difference at SETI.

You'd need to use an app_info file. That way you'd bypass BOINC checking the signatures on the files.

Strangely, the CUDA RT DLL is signed, but the FFT isn't. Seriously, does anyone know whether the mathematics here requires Fourier Transforms? If not, it's not worth chasing this.

And yes, I am using an app_info.xml

Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 1576
Credit: 5,600,136,851
RAC: 8,787,368
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11550 - Posted: 31 Jul 2009 | 14:51:48 UTC

1040687 completed and validated using the 2.3 DLLs. Took slightly longer than previous tasks on that box, but got slightly more credit - must have been an tougher job. Hourly rate exactly the same. Conclusion? It's not worth the hassle. Reverting to stock so I can get the new compiles as they come out.

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 11552 - Posted: 31 Jul 2009 | 16:33:37 UTC - in response to Message 11550.

We had already optimized the FFT our selves, so no much speedup will be seen using 2.3.
By the way, we have updated the application to cuda 2.2.

gdf

zpm
Avatar
Send message
Joined: 2 Mar 09
Posts: 159
Credit: 13,639,818
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11554 - Posted: 31 Jul 2009 | 17:17:30 UTC - in response to Message 11552.
Last modified: 31 Jul 2009 | 17:17:49 UTC

We had already optimized the FFT our selves, so no much speedup will be seen using 2.3.
By the way, we have updated the application to cuda 2.2.

gdf


thanks for updating, although last night i grabbed 40 seti cuda wu b/c i knew this change would happen today, and i just wanted to see what feedback was coming out from here... on the 6.67 app.

Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 1576
Credit: 5,600,136,851
RAC: 8,787,368
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11555 - Posted: 31 Jul 2009 | 17:47:56 UTC - in response to Message 11552.

We had already optimized the FFT our selves, so no much speedup will be seen using 2.3.

Ah well, experimental proof is always good to go alongside the theoretical conjecture.... :-)

By the way, we have updated the application to cuda 2.2.

gdf

And started distributing the 2.2 DLLs. Nice workround for the 'same name, different version' problem that BOINC handles so badly.

Sorry, I tried to unravel my app_info.xml without doing a full project reset - and failed. I've released the WU back into the wild for someone else to do.

And ended up with a reset project anyway. DCF 1.000000, <rsc_fpops_est>250000000000000 works out to 01:08:53 on this host, where it should be about 16 hours.

I could edit client_stats again, or you could - pretty please - put a more realistic estimate into the workunit generator?

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 11557 - Posted: 31 Jul 2009 | 23:02:12 UTC - in response to Message 11555.

We will start to increase by a factor 2. Then we need to wait a little.

gdf

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11559 - Posted: 1 Aug 2009 | 9:38:17 UTC
Last modified: 1 Aug 2009 | 9:40:28 UTC

I tried the Grid Wu's again last night when I seen the Application went to CUDA 2.2 but I'm still getting error's on at least 2 Box's with the 190.38 Drivers & 6.6.36 Client, same 2 Box's that have given me the most trouble since Updating to the 190.38 Drivers that didn't have any trouble before Upgrading.

Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 1576
Credit: 5,600,136,851
RAC: 8,787,368
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 11560 - Posted: 1 Aug 2009 | 9:45:20 UTC - in response to Message 11559.
Last modified: 1 Aug 2009 | 9:45:48 UTC

I tried the Grid Wu's again last night when I seen the Application went to CUDA 2.2 but I'm still getting error's on at least 2 Box's with the 190.38 Drivers & 6.6.36 Client, same 2 Box's that have given me the most trouble since Updating to the 190.38 Drivers that didn't have any trouble before Upgrading.

So which two boxes are they, and what graphics cards do they have installed?

Consensus seems to be settling down on 'G92-based cards run the new drivers OK, g200-based cards have problems'.

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11562 - Posted: 1 Aug 2009 | 11:42:30 UTC - in response to Message 11560.
Last modified: 1 Aug 2009 | 11:54:15 UTC

I tried the Grid Wu's again last night when I seen the Application went to CUDA 2.2 but I'm still getting error's on at least 2 Box's with the 190.38 Drivers & 6.6.36 Client, same 2 Box's that have given me the most trouble since Updating to the 190.38 Drivers that didn't have any trouble before Upgrading.

So which two boxes are they, and what graphics cards do they have installed?

Consensus seems to be settling down on 'G92-based cards run the new drivers OK, g200-based cards have problems'.


Right now it's Host ID 17909 & Host ID 38967 ... On the 17909 Host both 260 Cores were running @ 300/Core & 400/Memory when I checked it, On Host 38967 1 of the 260 Cores was running @ 300/Core & 400/Memory while the other Core was running 650/Core & 1100/Memory.

Both Box's had been running the Collatz Projects Wu's for the last 3-4 Days with no Error's or Re-setting of the Cores to Lower Speeds. I did a complete Uninstall of the NVIDIA Drivers again on both Box's, then Re-installed them to each 260 Card thru the Device Manager on both Box's, did the Riva Tuner Fix thingy on both Box's gave them both a Big Hug & Kiss to let them know they were still Loved ... hahaha ... and sent them on their way running the Grid Wu's again to see if that fixes them.

If they still spit out the errors then it's back to Collatz with them, so far the rest of my Pharm has behaved it's self and hasn't had any errors since last night.

What is the difference in the 'G92-based cards & the g200-based card ??? I just thought they all were 200 Series Cards. Funny thing is I have 4 Box's with 2 BFG GTX 260's each. 2 of the Box's seem to run the Wu's Okay but then the other 2 BFG GTX 260 Box's just spit out errors. All 8 Cards are the BFG 216 Shader OCX Version Type Cards ... Weird

NOTE: One thing I have noticed is that if one of the Cores drops or resets it's self to a lower running speed but the other one doesn't your still just going to get error's off both Cores. So the Speed drop must have an effect on both Cores for some reason or other.

I did a RMA on a BFG GTX 260 Card the same as the ones I have now just last week for the same reason of the Speed Drop. Since I got it's replacement back it's run okay, could just be something with the BFG Cards but I don't think so with all the other people reporting problems too.

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 11563 - Posted: 1 Aug 2009 | 12:01:03 UTC - in response to Message 11562.

it does not seem that you have successfully downgraded to 185.xx:
http://www.gpugrid.net/show_host_detail.php?hostid=17909

Your are having problems with the CUDA FFT library, which also seem to suggest that you are using the wrong driver. Please check.

Do you have the toolkit installed, if yes uninstall it. I am not sure if Windows would choose system installed libs rather than the one we provide.

gdf

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11564 - Posted: 1 Aug 2009 | 12:14:10 UTC - in response to Message 11563.

it does not seem that you have successfully downgraded to 185.xx:
http://www.gpugrid.net/show_host_detail.php?hostid=17909

Your are having problems with the CUDA FFT library, which also seem to suggest that you are using the wrong driver. Please check.

Do you have the toolkit installed, if yes uninstall it. I am not sure if Windows would choose system installed libs rather than the one we provide.

gdf


Hmmmmmm, but I tried that already (Downgrade to 185.18 I believe) & it din't work, I still got the errors, that and all my other Box's are running the 190.38 Drivers. I'll wait and see if they run okay now with the uninstall & reinstall I did with them an hour or so ago. If they still get the errors I'll try the 185.xx Drivers again.

No, I don't have the Toolkit installed on any of my Box's, did at first when I joined the Project but then un-installed them when I realized I didn't need them.

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11565 - Posted: 1 Aug 2009 | 12:33:36 UTC
Last modified: 1 Aug 2009 | 12:57:46 UTC

GDF, do you mean Downgrade to 185.xx Drivers or 186.xx Drivers, I think I tried the 186.18 Drivers but they didn't stop the Errors or the Cards from Dropping speed after just a few minutes of running sometimes, other times it took an hour or so before they dropped their speed.

PS: Crap, looks like something is wrong with one of the Box's already, 1 of the Wu's isn't Progressing according to BOINCView, will have to go out to the Garage & see what's up with it. Think I'll just go ahead and put the 185.xx Drivers in it & try them.

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11566 - Posted: 1 Aug 2009 | 13:19:49 UTC
Last modified: 1 Aug 2009 | 14:13:31 UTC

Looks like it did a Benchmark Test & upon Resuming the Wu's one of them erred out ... Host ID 17909 One Core & 300/Core & 400/Memory & One Core @ 650/Core & 1100/Memory ... Going to the 185.85 Drivers & see if they work or not ...

8/1/2009 8:14:24 AM Running CPU benchmarks
009 8:14:24 AM Suspending computation - running CPU benchmarks
8/1/2009 8:14:55 AM Benchmark results:
8/1/2009 8:14:55 AM Number of CPUs: 4
8/1/2009 8:14:55 AM 2275 floating point MIPS (Whetstone) per CPU
8/1/2009 8:14:55 AM 6944 integer MIPS (Dhrystone) per CPU
8/1/2009 8:14:56 AM Resuming computation
8/1/2009 8:29:00 AM AQUA@home Sending scheduler request: To fetch work.
8/1/2009 8:29:00 AM AQUA@home Requesting new tasks for CPU
8/1/2009 8:29:05 AM AQUA@home Scheduler request completed: got 0 new tasks
8/1/2009 8:29:05 AM AQUA@home Message from server: No work sent
8/1/2009 8:29:05 AM AQUA@home Message from server: (reached limit of 2 CPU tasks)
8/1/2009 9:10:05 AM GPUGRID Computation for task 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0 finished
8/1/2009 9:10:05 AM GPUGRID Output file 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0_1 for task 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0 absent
8/1/2009 9:10:05 AM GPUGRID Output file 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0_2 for task 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0 absent
8/1/2009 9:10:05 AM GPUGRID Output file 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0_3 for task 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0 absent
8/1/2009 9:10:06 AM Collatz Conjecture Restarting task collatz_1248745476_67361_1 using collatz version 110
8/1/2009 9:10:08 AM GPUGRID Started upload of 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0_0
8/1/2009 9:10:08 AM GPUGRID Started upload of 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0_4
8/1/2009 9:10:12 AM GPUGRID Finished upload of 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0_0
8/1/2009 9:11:37 AM GPUGRID Finished upload of 93-KASHIF_HIVPR_twomons_ba2-14-100-RND4610_0_4

PS: Have the 185'85 Drivers installed now, did find a couple of Odd Ball nv.dll's in the system 32 Directory, or at least i thought they were Odd Ball ones because the referred to an older Driver Version so I canned them before installing the 185.85 Drivers.

I also Aborted the running Grid Wu because I haven't had a Wu finish yet that was already running when one erred out, the remaining Wu would always error out eventually too so I figured 2 Fresh ones would be better. Am running both of them now with the 185.85 Drivers & 6.6.36 Client.

zpm
Avatar
Send message
Joined: 2 Mar 09
Posts: 159
Credit: 13,639,818
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11568 - Posted: 1 Aug 2009 | 14:10:12 UTC - in response to Message 11564.

it does not seem that you have successfully downgraded to 185.xx:
http://www.gpugrid.net/show_host_detail.php?hostid=17909

Your are having problems with the CUDA FFT library, which also seem to suggest that you are using the wrong driver. Please check.

Do you have the toolkit installed, if yes uninstall it. I am not sure if Windows would choose system installed libs rather than the one we provide.

gdf


Hmmmmmm, but I tried that already (Downgrade to 185.18 I believe) & it din't work, I still got the errors, that and all my other Box's are running the 190.38 Drivers. I'll wait and see if they run okay now with the uninstall & reinstall I did with them an hour or so ago. If they still get the errors I'll try the 185.xx Drivers again.

No, I don't have the Toolkit installed on any of my Box's, did at first when I joined the Project but then un-installed them when I realized I didn't need them.



and people wonder why some of us install the toolkit....... here's your asnwer....

i install the toolkit each time it's updated... i rarely have an issue...

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11569 - Posted: 1 Aug 2009 | 14:16:31 UTC - in response to Message 11568.
Last modified: 1 Aug 2009 | 14:19:47 UTC

it does not seem that you have successfully downgraded to 185.xx:
http://www.gpugrid.net/show_host_detail.php?hostid=17909

Your are having problems with the CUDA FFT library, which also seem to suggest that you are using the wrong driver. Please check.

Do you have the toolkit installed, if yes uninstall it. I am not sure if Windows would choose system installed libs rather than the one we provide.

gdf


Hmmmmmm, but I tried that already (Downgrade to 185.18 I believe) & it din't work, I still got the errors, that and all my other Box's are running the 190.38 Drivers. I'll wait and see if they run okay now with the uninstall & reinstall I did with them an hour or so ago. If they still get the errors I'll try the 185.xx Drivers again.

No, I don't have the Toolkit installed on any of my Box's, did at first when I joined the Project but then un-installed them when I realized I didn't need them.



and people wonder why some of us install the toolkit....... here's your asnwer....

i install the toolkit each time it's updated... i rarely have an issue...


Oh Please zpm, spare me the drama, I've never had the Toolkit installed and never had any problems either ... If you have something to add that's actually helpful please post other wise take your redundant spam some place else ... None of my other Box's have the Toolkit installed and they run fine here, it's just something with these 2 Box's and so far this morning just 1 Box.

zpm
Avatar
Send message
Joined: 2 Mar 09
Posts: 159
Credit: 13,639,818
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11570 - Posted: 1 Aug 2009 | 14:29:43 UTC - in response to Message 11569.
Last modified: 1 Aug 2009 | 14:32:11 UTC

what i was trying to say PB; is that it can't hurt to have it installed(unless your just; no space available).


btw, if you think that post was meant for you as i hit the quote button on your post and didn't realize it... it wasn't meant for you, it was just meant to stat the fact that, some extra's tools in the box can't hurt...

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11571 - Posted: 1 Aug 2009 | 14:37:38 UTC - in response to Message 11570.
Last modified: 1 Aug 2009 | 14:43:57 UTC

what i was trying to say PB; is that it can't hurt to have it installed(unless your just; no space available).


btw, if you think that post was meant for you as i hit the quote button on your post and didn't realize it... it wasn't meant for you, it was just meant to stat the fact that, some extra's tools in the box can't hurt...


Then say what you mean instead of just smugly throwing something out there that makes me look like an idiot for not having something installed that you do.

I'll try the toolkit later if going back to the 185.85 Drivers don't work on that Box ... Never have had to have it installed but as you say it might not hurt if all else fails ...

PS: Crap again, one just errored out on that Box so the 185.85 Driver rollback didn't work & I notice one looks like it's stalled on the other Box. The last time I noticed one stalled it was erred out when I checked the Box.

I don't have time to mess with them anymore today and I'm just paying for electricity that's getting nothing done so I'm just going to put them back on Collatz until this afternoon when I can mess with them a little again. I'll try the Toolkit then ...

zpm
Avatar
Send message
Joined: 2 Mar 09
Posts: 159
Credit: 13,639,818
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11572 - Posted: 1 Aug 2009 | 14:48:21 UTC - in response to Message 11571.

poorboy, i hate to say this but when i ran aqua cuda, my gpugrid wu's errored out until the cuda aqua was off my computer(boinc)... i doubt that collatz is affecting the gpugrid wu's but who knows.....

something else too look into...

if all else fails, detach, and reattach...

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11575 - Posted: 1 Aug 2009 | 15:42:56 UTC - in response to Message 11572.

poorboy, i hate to say this but when i ran aqua cuda, my gpugrid wu's errored out until the cuda aqua was off my computer(boinc)... i doubt that collatz is affecting the gpugrid wu's but who knows.....

something else too look into...

if all else fails, detach, and reattach...


I'm running the AQUA MT not AQUA CUDA @ Collatz CUDA Wu's on all of my NVIDIA Box's so you would think if it affected 1 or 2 Box's it would affect all of them.

I did install the CUDA 2.2 Toolkit on the Box with the 185.85 Drivers and CUDA 2.3 Toolkit on the Box with the 190.38 Drivers so we'll see if that works or not.

I've used up my quota on the 17909 Box for the Day & only have 1 started Wu left on that until tonight, the other Box has 2 Wu's on it & their both running now. Will check them later & see what happened with them.

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11578 - Posted: 1 Aug 2009 | 20:05:35 UTC

Both Box's have erred out their remaining Wu's using the 185.85 & 190.38 Drivers with & without the CUDA 2.2 & 2.3 Toolkit installed. About the only thing left to those to Box's is pull each card individually & see if it will run the Wu's or not on it's own. if it will then that particular Card probably good but if it won't then the Card is bad I would suspect.

I'll try to RMA them then if that's the case, I would think BFG's gonna get tired sooner or later of RMA'ing Video Cards just because they won't run the CUDA Applications here & will refuse to RMA them at some point.

zpm
Avatar
Send message
Joined: 2 Mar 09
Posts: 159
Credit: 13,639,818
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11579 - Posted: 1 Aug 2009 | 21:34:12 UTC - in response to Message 11578.
Last modified: 1 Aug 2009 | 21:34:59 UTC

unless they bluescreen and die on you, the cards are ok....

what cards are they.... i'm sure your not the only one with a bunch of cards throwing you errors...

best thing to do, is to go to nvidia forums for cuda and see if any person with same card and all is having a problem doing the same things..

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 11580 - Posted: 1 Aug 2009 | 22:35:57 UTC - in response to Message 11579.

PoorBoy,
your error is quite odd. It always fails in the fft calls.
Try to look in the system for cufft.dll. You should have only the one distributed by gpugrid.

gdf

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11582 - Posted: 1 Aug 2009 | 23:09:11 UTC - in response to Message 11580.
Last modified: 1 Aug 2009 | 23:46:59 UTC

PoorBoy,
your error is quite odd. It always fails in the fft calls.
Try to look in the system for cufft.dll. You should have only the one distributed by gpugrid.

gdf


First off I did some checking around the NVIDIA Forums & couldn't find anything oertaining to my Problem. Found a little stuff on Linux & BOINC But nothing on Windows & BOINC.

Host ID 17909 has 2 cufft.dll's:

1 in C:\CUDA\bin < CUDA Toolkit I think
1 in D:\Program Files\BOINC\projects\www.gpu.grid.net
None in the slot's as no Wu's were running

Host ID 38967 has 4 cufft.dll's:

1 in C:\CUDA\bin < CUDA Toolkit I think
1 in D:\Program Files\BOINC\projects\www.gpu.grid.net
1 in D:\Program Files\BOINC\slot3
1 in D:\Program Files\BOINC\slot4
2 Wu's running on that Box 42% & 39% done after about 3.5 Hours of running, haven't erred out yet using 185.85 Drivers with the CUDA 2.2 Toolkit installed

zpm
Avatar
Send message
Joined: 2 Mar 09
Posts: 159
Credit: 13,639,818
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 11583 - Posted: 1 Aug 2009 | 23:48:20 UTC - in response to Message 11582.
Last modified: 1 Aug 2009 | 23:53:15 UTC

best i can say, if nothing else works... revert to 186.38.... and wait until next month...

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11584 - Posted: 2 Aug 2009 | 0:11:06 UTC - in response to Message 11583.

best i can say, if nothing else works... revert to 186.38.... and wait until next month...


I've tried the 186.18's already and they didn't work, not sure there even is a 186.38 Driver. I'm downloading some 190.40 Drivers now & will try them tonight or tomorrow morning on those 2 Box's. The other Box's are runing okay so I'm not messing with them unless I have to.

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11591 - Posted: 2 Aug 2009 | 8:46:44 UTC
Last modified: 2 Aug 2009 | 9:02:19 UTC

Tried the 190.40 BETA Drivers over night on just 1 of the Dual GTX 260 Box's, the good news is after about 7+ hours of running neither Wu had erred out, the bad news is one of the Wu quit progressing some time during the night around 43% completion while the other Wu was @ 83% completion.

Both 260 Cores still showed full speed with GPU-Z, stopping & restarting BOINC got the stalled Wu running again but I could tell @ only 1/2 speed or less. Checking the Cores again showed 1 Core to be @ the all to familiar 300/Core & 400/Memory Speeds. So I rebooted the Box & when BOINC started again both Cores were at full speed again.

I'll let them run & see what happens with them from here. I also started up the other Box that's been trouble with the 190.40 BETA Drivers to see what it's going to do with them.

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11616 - Posted: 2 Aug 2009 | 20:18:31 UTC

Well 3 of the 4 Wu's on the 2 Box's ended up with error's, 1 of them was within a few % Points of finishing when it erred out. So I'm done for awhile running the Grid Wu's until somebody comes up with a fix. The Riva Fix may work on some Box's but it doesn't work on any of mine so thats out.

I found 1 Wu on my Dual 275 Box this Morning that had been running for 18+ hr's according to BOINC Manager's CPU Time so it probably has been runing much longer than that. It was at 34% done, 1 Core was at the dreaded 300/Core 400/Memory, I just aborted it and set the box to NNW.

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11617 - Posted: 2 Aug 2009 | 22:58:00 UTC

I made a Post over at the NVIDIA CUDA Forum, maybe somebody has a fix or suggestions over there I can try ...

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 11622 - Posted: 3 Aug 2009 | 5:32:50 UTC

All WUs exiting with "output file ... is null" error. BOINC 6.6.38, 190.38 nVidia driver, Win 7 Ultimate x64.
____________
From Siberia with love!

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11639 - Posted: 3 Aug 2009 | 20:37:40 UTC
Last modified: 3 Aug 2009 | 20:42:55 UTC

I have talked to BFG & they are in agreement something is wrong with the Cards if they Reset themselves to a lower clock speed. They are willing to RMA all of my Cards that do that too, I have nothing but good stuff to say about BFG, they have RMA'ed several Cards for me already without any Questions, even sent me a Pre-Paid Shipping Label on the last one. The BFG Cards may cost a little more but to me their well worth it for their Warranty Service.

BFG is also going to forward my E-Mail to them to NVIDIA & see if they can't get NVIDIA to include Windows XP with the Prefer Maximum Performance Setting in the NVIDIA Control Panel Settings that Vista & on up got with the Recent releases. Maybe PoorBoy with BFG's help can move a Giant Company to do something for him & all fellow XP users that run NVIDIA Cards ... One can hope anyway ... B)

shadowwind
Send message
Joined: 4 Jul 09
Posts: 3
Credit: 6,448,856
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 11642 - Posted: 3 Aug 2009 | 22:31:40 UTC - in response to Message 11639.

that is good news glad to read some thing good from you i was having the same problems i got one XFX gtx 260 that went back last Tuesday i couldn't get that card to work i did all the things you did. good luck and i hope you get things going as soon as possible.

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 313,909,798
RAC: 486,126
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 11644 - Posted: 4 Aug 2009 | 0:11:24 UTC - in response to Message 11642.

that is good news glad to read some thing good from you i was having the same problems i got one XFX gtx 260 that went back last Tuesday i couldn't get that card to work i did all the things you did. good luck and i hope you get things going as soon as possible.


All the GTX Cards I have are running, just not running the Grid Project, they run the Collatz, Seti, AQUA CUDA Wu's just fine so it's not like their out of commission.

RMA'ing them will be the last resort if NVIDIA doesn't release some Drivers that fix the problem, until then I just won't be able to run the Grid Wu's with certain Cards I have.

Post to thread

Message boards : Graphics cards (GPUs) : Nvidia 190.38, Cuda 2.3

//