Advanced search

Message boards : Number crunching : A problem with "Full atom molecular dynamics 6.43"

Author Message
Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 2014 - Posted: 2 Sep 2008 | 12:03:30 UTC

Hi all!

A lot of WUs received by my BOINC client are unable to be calculated till the end. BOINC says:

02.09.2008 18:44:19|PS3GRID|Restarting task Orl8543-GPUTEST2-2-10-acemd_0 using acemd version 643
02.09.2008 18:44:22|lhcathome|Started upload of w3_lhc_symmetric-q1_8__38__s__64.31_59.32__12_14__5__45_1_sixvf_boinc8599_1_0
02.09.2008 18:44:23|PS3GRID|Computation for task Orl8543-GPUTEST2-2-10-acemd_0 finished
02.09.2008 18:44:23|PS3GRID|Output file Orl8543-GPUTEST2-2-10-acemd_0_1 for task Orl8543-GPUTEST2-2-10-acemd_0 absent
02.09.2008 18:44:23|PS3GRID|Output file Orl8543-GPUTEST2-2-10-acemd_0_2 for task Orl8543-GPUTEST2-2-10-acemd_0 absent
02.09.2008 18:44:23|PS3GRID|Output file Orl8543-GPUTEST2-2-10-acemd_0_3 for task Orl8543-GPUTEST2-2-10-acemd_0 absent
My GPU is 8600GT, and untill now not even one WU was finished well. Why? And how to deal with it?

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 2016 - Posted: 2 Sep 2008 | 12:36:27 UTC - in response to Message 2014.
Last modified: 2 Sep 2008 | 12:36:47 UTC

Hi all!

A lot of WUs received by my BOINC client are unable to be calculated till the end. BOINC says:

02.09.2008 18:44:19|PS3GRID|Restarting task Orl8543-GPUTEST2-2-10-acemd_0 using acemd version 643
02.09.2008 18:44:22|lhcathome|Started upload of w3_lhc_symmetric-q1_8__38__s__64.31_59.32__12_14__5__45_1_sixvf_boinc8599_1_0
02.09.2008 18:44:23|PS3GRID|Computation for task Orl8543-GPUTEST2-2-10-acemd_0 finished
02.09.2008 18:44:23|PS3GRID|Output file Orl8543-GPUTEST2-2-10-acemd_0_1 for task Orl8543-GPUTEST2-2-10-acemd_0 absent
02.09.2008 18:44:23|PS3GRID|Output file Orl8543-GPUTEST2-2-10-acemd_0_2 for task Orl8543-GPUTEST2-2-10-acemd_0 absent
02.09.2008 18:44:23|PS3GRID|Output file Orl8543-GPUTEST2-2-10-acemd_0_3 for task Orl8543-GPUTEST2-2-10-acemd_0 absent
My GPU is 8600GT, and untill now not even one WU was finished well. Why? And how to deal with it?



Can you specify BOINC client version, operating system and Nvidia driver installed?

thanks, gdf

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 2020 - Posted: 2 Sep 2008 | 13:18:04 UTC

Sure!

02.09.2008 16:22:52||Starting BOINC client version 6.3.10 for windows_x86_64
02.09.2008 16:22:52||This a development version of BOINC and may not function properly
02.09.2008 16:22:52||log flags: task, file_xfer, sched_ops
02.09.2008 16:22:52||Libraries: libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3
02.09.2008 16:22:52||Data directory: C:\Documents and Settings\All Users\Application Data\BOINC
02.09.2008 16:22:52||Running under account bla-bla-bla
02.09.2008 16:22:54||Processor: 2 AuthenticAMD AMD Athlon(tm) 64 X2 Dual Core Processor 5200+ [AMD64 Family 15 Model 67 Stepping 2]
02.09.2008 16:22:54||Processor features: fpu tsc pae nx sse sse2
02.09.2008 16:22:54||OS: Microsoft Windows XP: Professional x64 Editon, Service Pack 2, (05.02.3790.00)
02.09.2008 16:22:54||Memory: 2.00 GB physical, 3.87 GB virtual
02.09.2008 16:22:54||Disk: 19.53 GB total, 5.20 GB free
02.09.2008 16:22:54||Local time is UTC +7 hours
02.09.2008 16:22:54||Not using a proxy
02.09.2008 16:23:02||CUDA devices found
02.09.2008 16:23:02||Coprocessor: GeForce 8600 GT (1)

ForceWare version: 177.84 (as nVidia COntrol Panel says :-)).

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 2034 - Posted: 2 Sep 2008 | 18:58:52 UTC - in response to Message 2020.

Sure!

02.09.2008 16:22:52||Starting BOINC client version 6.3.10 for windows_x86_64
02.09.2008 16:22:52||This a development version of BOINC and may not function properly
02.09.2008 16:22:52||log flags: task, file_xfer, sched_ops
02.09.2008 16:22:52||Libraries: libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3
02.09.2008 16:22:52||Data directory: C:\Documents and Settings\All Users\Application Data\BOINC
02.09.2008 16:22:52||Running under account bla-bla-bla
02.09.2008 16:22:54||Processor: 2 AuthenticAMD AMD Athlon(tm) 64 X2 Dual Core Processor 5200+ [AMD64 Family 15 Model 67 Stepping 2]
02.09.2008 16:22:54||Processor features: fpu tsc pae nx sse sse2
02.09.2008 16:22:54||OS: Microsoft Windows XP: Professional x64 Editon, Service Pack 2, (05.02.3790.00)
02.09.2008 16:22:54||Memory: 2.00 GB physical, 3.87 GB virtual
02.09.2008 16:22:54||Disk: 19.53 GB total, 5.20 GB free
02.09.2008 16:22:54||Local time is UTC +7 hours
02.09.2008 16:22:54||Not using a proxy
02.09.2008 16:23:02||CUDA devices found
02.09.2008 16:23:02||Coprocessor: GeForce 8600 GT (1)

ForceWare version: 177.84 (as nVidia COntrol Panel says :-)).




Somehow your windows system is using too much memory and the application runs out of memory. Try to simply the Windows effects of the desktop and restart the machine.
In principle 256 Mb are sufficient.

gdf

GPUGRID Role account
Send message
Joined: 15 Feb 07
Posts: 134
Credit: 1,349,535,983
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 2038 - Posted: 2 Sep 2008 | 20:07:57 UTC - in response to Message 2034.


Somehow your windows system is using too much memory and the application runs out of memory.


If you have a game running when BOINC starts the GPUGRID application, it is quite likely that it fail because the game has taken all of the available device memory.

MJH

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 2050 - Posted: 3 Sep 2008 | 5:41:10 UTC

It may looks strange, but my machine doing nothing else when it processing PS3GRID WUs. I left my flat for several days, so nobody (may be, spirits?) could use this PC for any need. From another hand, it always shows that about 1.3 Gb of RAM is free.

May be somehow 256 MB of video RAM is not enough?..

May be it's better to stop calculations on my GPU untill I will upgrade it? But I am ready to participate, and my PC often is absolutely free...

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 2058 - Posted: 3 Sep 2008 | 9:30:40 UTC

Another one case:

http://www.gpugrid.net/result.php?resultid=50240

It takes 5000+ secs and then crashed.

p.s. I thinf to buy GTX260 for PS3GRID computation, but it seems the problem occures on it too sometimes...

Rabinovitch
Avatar
Send message
Joined: 25 Aug 08
Posts: 143
Credit: 64,937,578
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 2229 - Posted: 10 Sep 2008 | 4:49:57 UTC

Well, now I'm crunching on GTX260. But there is still a fact: 0ne CPU core is totally busy during a PS3GRID WU processing. Someone told somwhere here, in project forums, that should not be so - "CPU is only need to organize work".

Please tell me someone, why Full atom molecular dynamics 6.43 application using CPU so much?

Profile koschi
Avatar
Send message
Joined: 14 Aug 08
Posts: 124
Credit: 486,829,198
RAC: 1,362,608
Level
Gln
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 2230 - Posted: 10 Sep 2008 | 4:59:48 UTC

Actually that core is not doing any real work, just polling the GPU app about its status, as I understood.

This issue will be addressed in future BOINC versions, so that we can use those core again for other projects.

Right now you have to cope with that ;-) This project (the GPUGRID part) is bleeding-edge and hence not perfect yet...

Barraud Denis
Avatar
Send message
Joined: 2 Sep 08
Posts: 15
Credit: 36,207,656
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 2303 - Posted: 13 Sep 2008 | 7:39:44 UTC

I don't succed to crunch ps3grid unit. my pc: Q9550 / 2 Go Of DDR2-1066, Vista 64bits, HD4850 ... I sugest you to rewrite deviceQuery.cu completely and recompile .

13-Sep-2008 01:48:54 [---] Starting BOINC client version 6.3.10 for windows_x86_64
13-Sep-2008 01:48:54 [---] This a development version of BOINC and may not function properly
13-Sep-2008 01:48:54 [---] log flags: task, file_xfer, sched_ops
13-Sep-2008 01:48:54 [---] Libraries: libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3
13-Sep-2008 01:48:54 [---] Data directory: E:\ProgramData\BOINC
13-Sep-2008 01:48:54 [---] Running under account BarraudDen
13-Sep-2008 01:48:54 [---] Processor: 4 GenuineIntel Intel(R) Core(TM)2 Quad CPU Q9550 @ 2.83GHz [Intel64 Family 6 Model 23 Stepping 7]
13-Sep-2008 01:48:54 [---] Processor features: fpu tsc pae nx sse sse2 pni
13-Sep-2008 01:48:54 [---] OS: Microsoft Windows Vista: Business x64 Editon, Service Pack 1, (06.00.6001.00)
13-Sep-2008 01:48:54 [---] Memory: 2.00 GB physical, 8.53 GB virtual
13-Sep-2008 01:48:54 [---] Disk: 698.63 GB total, 667.54 GB free
13-Sep-2008 01:48:54 [---] Local time is UTC +2 hours
13-Sep-2008 01:48:54 [---] Not using a proxy
13-Sep-2008 01:48:54 [---] CUDA devices found
13-Sep-2008 01:48:54 [---] Coprocessor: Device Emulation (CPU) (1)
13-Sep-2008 01:48:54 [---] Version change (6.2.18 -> 6.3.10)
-----
04:38:56 [PS3GRID] Starting task Azn7836-GPUTEST2-5-10-acemd_0 using acemd version 643
13-Sep-2008 04:38:57 [PS3GRID] Computation for task Azn7836-GPUTEST2-5-10-acemd_0 finished
13-Sep-2008 04:38:57 [PS3GRID] Output file Azn7836-GPUTEST2-5-10-acemd_0_1 for task Azn7836-GPUTEST2-5-10-acemd_0 absent
13-Sep-2008 04:38:57 [PS3GRID] Output file Azn7836-GPUTEST2-5-10-acemd_0_2 for task Azn7836-GPUTEST2-5-10-acemd_0 absent
13-Sep-2008 04:38:57 [PS3GRID] Output file Azn7836-GPUTEST2-5-10-acemd_0_3 for task Azn7836-GPUTEST2-5-10-acemd_0 absent
-----
13-Sep-2008 04:38:58 [PS3GRID] Computation for task LFk6908-GPUTEST3-1-10-acemd_1 finished
13-Sep-2008 04:38:58 [PS3GRID] Output file LFk6908-GPUTEST3-1-10-acemd_1_1 for task LFk6908-GPUTEST3-1-10-acemd_1 absent
13-Sep-2008 04:38:58 [PS3GRID] Output file LFk6908-GPUTEST3-1-10-acemd_1_2 for task LFk6908-GPUTEST3-1-10-acemd_1 absent
13-Sep-2008 04:38:58 [PS3GRID] Output file LFk6908-GPUTEST3-1-10-acemd_1_3 for task LFk6908-GPUTEST3-1-10-acemd_1 absent
-----

every time in task result:

core_client_version>6.3.10</core_client_version>
<![CDATA[
<message>
Fonction incorrecte. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
# Using CUDA device 0
Cuda error in file 'deviceQuery.cu' in line 59 : feature is not yet implemented.

</stderr_txt>
]]>

Wolfram1
Send message
Joined: 24 Aug 08
Posts: 45
Credit: 3,431,862
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwatwat
Message 2304 - Posted: 13 Sep 2008 | 8:50:48 UTC - in response to Message 2303.

I don't succed to crunch ps3grid unit. my pc: Q9550 / 2 Go Of DDR2-1066, Vista 64bits, HD4850 ... I sugest you to rewrite deviceQuery.cu completely and recompile .


13-Sep-2008 01:48:54 [---] CUDA devices found
13-Sep-2008 01:48:54 [---] Coprocessor: Device Emulation (CPU) (1)




Your problem is in the second line. There must be the card number like this:

09.09.2008 23:37:57||Coprocessor: GeForce 9800 GTX/9800 GTX+ (1)

During installation of Boinc 6.3.10 you must *not* install in protect mode. You mist click off (i think) the second box)

Profile UBT - NaRyan
Avatar
Send message
Joined: 16 Jul 08
Posts: 68
Credit: 1,242,980
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 2305 - Posted: 13 Sep 2008 | 9:31:22 UTC - in response to Message 2304.
Last modified: 13 Sep 2008 | 9:31:51 UTC

PS3GRID won't work for him.
"Vista 64bits, HD4850..."(That's an ATI GPU) :(
____________

Down with the Kredit Kops!!!

Cyborg
Send message
Joined: 22 Sep 08
Posts: 2
Credit: 7,413,924
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwatwat
Message 2579 - Posted: 24 Sep 2008 | 6:11:06 UTC

Hi all,
I just joined the project and 2 of the 3 WUs generated a Computation error:

24.09.2008 07:46:29|PS3GRID|Computation for task Cl18636-GPUTEST3-4-10-acemd_0 finished
24.09.2008 07:46:29|PS3GRID|Output file Cl18636-GPUTEST3-4-10-acemd_0_1 for task Cl18636-GPUTEST3-4-10-acemd_0 absent
24.09.2008 07:46:29|PS3GRID|Output file Cl18636-GPUTEST3-4-10-acemd_0_2 for task Cl18636-GPUTEST3-4-10-acemd_0 absent
24.09.2008 07:46:29|PS3GRID|Output file Cl18636-GPUTEST3-4-10-acemd_0_3 for task Cl18636-GPUTEST3-4-10-acemd_0 absent
24.09.2008 07:46:31|PS3GRID|Started upload of Cl18636-GPUTEST3-4-10-acemd_0_0
24.09.2008 07:46:35|PS3GRID|Finished upload of Cl18636-GPUTEST3-4-10-acemd_0_0


23.09.2008 20:59:59||Starting BOINC client version 6.3.10 for windows_x86_64
23.09.2008 20:59:59||log flags: task, file_xfer, sched_ops
23.09.2008 20:59:59||Libraries: libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3
23.09.2008 20:59:59||Data directory: C:\Documents and Settings\All Users\Application Data\BOINC
23.09.2008 20:59:59||Running under account ...
23.09.2008 20:59:59|SETI@home|Found app_info.xml; using anonymous platform
23.09.2008 20:59:59||Processor: 4 GenuineIntel Intel(R) Core(TM)2 Quad CPU Q9450 @ 2.66GHz [EM64T Family 6 Model 23 Stepping 7]
23.09.2008 20:59:59||Processor features: fpu tsc pae nx sse sse2
23.09.2008 20:59:59||OS: Microsoft Windows XP: Professional x64 Editon, Service Pack 2, (05.02.3790.00)
23.09.2008 20:59:59||Memory: 8.00 GB physical, 9.58 GB virtual
23.09.2008 20:59:59||Disk: 97.65 GB total, 75.21 GB free
23.09.2008 20:59:59||Local time is UTC +2 hours
23.09.2008 20:59:59||Not using a proxy
23.09.2008 21:00:00||CUDA devices found
23.09.2008 21:00:00||Coprocessor: GeForce 9600 GT (1)

nVidia driver is 177.84

<core_client_version>6.3.10</core_client_version>
<![CDATA[
<message>
Unzulässige Funktion. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
# Using CUDA device 0
# Device 0: "GeForce 9600 GT"
# Clock rate: 1600000 kilohertz
MDIO ERROR: cannot open file "restart.coor"
# Using CUDA device 0
# Device 0: "GeForce 9600 GT"
# Clock rate: 1600000 kilohertz
# Using CUDA device 0
# Device 0: "GeForce 9600 GT"
# Clock rate: 1600000 kilohertz
Cuda error: Kernel [kick_drift_kernel] failed in file 'step.cu' in line 46 : unspecified launch failure.

</stderr_txt>
]]>

Can someone tell me what the problem is?

C

Cyborg
Send message
Joined: 22 Sep 08
Posts: 2
Credit: 7,413,924
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwatwat
Message 2642 - Posted: 26 Sep 2008 | 20:31:15 UTC

...and another one:

http://www.gpugrid.net/result.php?resultid=65379

C

Post to thread

Message boards : Number crunching : A problem with "Full atom molecular dynamics 6.43"

//