I have n noticed that my NVIDEA GPU tasks for the Moo! Wrapper project are not running on my dual GeForce GTX 550 Ti Windows 7 Premium x64 computer. All CUDA tasks seemed to have stopped running when I upgraded to BOINC V 7.0.28. Most of the other projects just didn't download any CUDA tasks, but Moo! Wrapper did. Someone on the BOINC forum suggested that I roll back to BOINC V 7.0.25 and this has allowed some of the other CUDA projects to download and run on my machine. I saw a similar problem with someone running dual-ATI cards on this message board. The solution there was some sort of configuration file, but it was specific to ATI cards. Is there something similar for NVIDEA cars?

Here's my startup from the BOINC event log:


9/15/2012 10:54:14 PM | | No config file found - using defaults
9/15/2012 10:54:14 PM | | Starting BOINC client version 7.0.25 for windows_x86_64
9/15/2012 10:54:14 PM | | log flags: file_xfer, sched_ops, task
9/15/2012 10:54:14 PM | | Libraries: libcurl/7.21.6 OpenSSL/1.0.0d zlib/1.2.5
9/15/2012 10:54:14 PM | | Data directory: C:\ProgramData\BOINC
9/15/2012 10:54:14 PM | | Running under account Dude
9/15/2012 10:54:14 PM | | Processor: 4 AuthenticAMD AMD FX(tm)-4100 Quad-Core Processor [Family 21 Model 1 Stepping 2]
9/15/2012 10:54:14 PM | | Processor: 2.00 MB cache
9/15/2012 10:54:14 PM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 htt pni ssse3 cx16 sse4_1 sse4_2 syscall nx lm svm sse4a osvw ibs xop skinit wdt lwp fma4 page1gb rdtscp
9/15/2012 10:54:14 PM | | OS: Microsoft Windows 7: Home Premium x64 Edition, Service Pack 1, (06.01.7601.00)
9/15/2012 10:54:14 PM | | Memory: 16.00 GB physical, 31.99 GB virtual
9/15/2012 10:54:14 PM | | Disk: 494.73 GB total, 375.00 GB free
9/15/2012 10:54:14 PM | | Local time is UTC -4 hours
9/15/2012 10:54:14 PM | | VirtualBox version: 4.1.8
9/15/2012 10:54:14 PM | | NVIDIA GPU 0: GeForce GTX 550 Ti (driver version 301.42, CUDA version 4.20, compute capability 2.1, 1024MB, 600MB available, 703 GFLOPS peak)
9/15/2012 10:54:14 PM | | NVIDIA GPU 1: GeForce GTX 550 Ti (driver version 301.42, CUDA version 4.20, compute capability 2.1, 1024MB, 600MB available, 703 GFLOPS peak)
9/15/2012 10:54:14 PM | | OpenCL: NVIDIA GPU 0: GeForce GTX 550 Ti (driver version 301.42, device version OpenCL 1.1 CUDA, 1024MB, 600MB available)
9/15/2012 10:54:14 PM | | OpenCL: NVIDIA GPU 1: GeForce GTX 550 Ti (driver version 301.42, device version OpenCL 1.1 CUDA, 1024MB, 600MB available)
9/15/2012 10:54:14 PM | | Version change (7.0.28 -> 7.0.25)
9/15/2012 10:54:14 PM | Constellation | URL http://aerospaceresearch.net/constellation/; Computer ID 7006; resource share 500
9/15/2012 10:54:14 PM | rosetta@home | URL http://boinc.bakerlab.org/rosetta/; Computer ID 1503442; resource share 150
9/15/2012 10:54:14 PM | DistrRTgen | URL http://boinc.freerainbowtables.com/distrrtgen/; Computer ID 31325; resource share 500
9/15/2012 10:54:14 PM | Poem@Home | URL http://boinc.fzk.de/poem/; Computer ID 110049; resource share 100
9/15/2012 10:54:14 PM | Leiden Classical | URL http://boinc.gorlaeus.net/; Computer ID 92770; resource share 100
9/15/2012 10:54:14 PM | Collatz Conjecture | URL http://boinc.thesonntags.com/collatz/; Computer ID 90139; resource share 200
9/15/2012 10:54:14 PM | The Lattice Project | URL http://boinc.umiacs.umd.edu/; Computer ID 73419; resource share 100
9/15/2012 10:54:14 PM | boincsimap | URL http://boincsimap.org/boincsimap/; Computer ID 219448; resource share 100
9/15/2012 10:54:14 PM | BURP | URL http://burp.renderfarming.net/; Computer ID 52919; resource share 500
9/15/2012 10:54:14 PM | climateprediction.net | URL http://climateprediction.net/; Computer ID 1185692; resource share 200
9/15/2012 10:54:14 PM | Docking | URL http://docking.cis.udel.edu/; Computer ID 107264; resource share 100
9/15/2012 10:54:14 PM | Einstein@Home | URL http://einstein.phys.uwm.edu/; Computer ID 4309125; resource share 200
9/15/2012 10:54:14 PM | Test4Theory@Home | URL http://lhcathome2.cern.ch/test4theory/; Computer ID 14461; resource share 500
9/15/2012 10:54:14 PM | LHC@home 1.0 | URL http://lhcathomeclassic.cern.ch/sixtrack/; Computer ID 9948039; resource share 100
9/15/2012 10:54:14 PM | Milkyway@Home | URL http://milkyway.cs.rpi.edu/milkyway/; Computer ID 355680; resource share 200
9/15/2012 10:54:14 PM | MindModeling@Beta | URL http://MindModeling.org/beta/; Computer ID 19162; resource share 100
9/15/2012 10:54:14 PM | Moo! Wrapper | URL http://moowrap.net/; Computer ID 7910; resource share 200
9/15/2012 10:54:14 PM | orbit@home | URL http://orbit.psi.edu/oah/; Computer ID 82049; resource share 100
9/15/2012 10:54:14 PM | Pirates@Home | URL http://pirates.spy-hill.net/; Computer ID 71799; resource share 100
9/15/2012 10:54:14 PM | QMC@HOME | URL http://qah.uni-muenster.de/; Computer ID 203296; resource share 100
9/15/2012 10:54:14 PM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 6293163; resource share 1000
9/15/2012 10:54:14 PM | SETI@home Beta Test | URL http://setiweb.ssl.berkeley.edu/beta/; Computer ID 55373; resource share 100
9/15/2012 10:54:14 PM | Spinhenge@home | URL http://spin.fh-bielefeld.de/; Computer ID 220826; resource share 100
9/15/2012 10:54:14 PM | Cosmology@Home | URL http://www.cosmologyathome.org/; Computer ID 151072; resource share 100
9/15/2012 10:54:14 PM | Enigma@Home | URL http://www.enigmaathome.net/; Computer ID 71412; resource share 100
9/15/2012 10:54:14 PM | FreeHAL@home | URL http://www.freehal.net/freehal_at_home/; Computer ID 72013; resource share 100
9/15/2012 10:54:14 PM | GPUGRID | URL http://www.gpugrid.net/; Computer ID 115985; resource share 200
9/15/2012 10:54:14 PM | PrimeGrid | URL http://www.primegrid.com/; Computer ID 240499; resource share 100
9/15/2012 10:54:14 PM | yoyo@home | URL http://www.rechenkraft.net/yoyo/; Computer ID 63254; resource share 100
9/15/2012 10:54:14 PM | uFluids | URL http://www.ufluids.net/; Computer ID 135801; resource share 100
9/15/2012 10:54:14 PM | World Community Grid | URL http://www.worldcommunitygrid.org/; Computer ID 1830670; resource share 200
9/15/2012 10:54:14 PM | Moo! Wrapper | General prefs: from Moo! Wrapper (last modified 09-Aug-2012 19:33:33)
9/15/2012 10:54:14 PM | Moo! Wrapper | Host location: none
9/15/2012 10:54:14 PM | Moo! Wrapper | General prefs: using your defaults
9/15/2012 10:54:14 PM | | Reading preferences override file
9/15/2012 10:54:14 PM | | Preferences:
9/15/2012 10:54:14 PM | | max memory usage when active: 12285.93MB
9/15/2012 10:54:14 PM | | max memory usage when idle: 16381.24MB
9/15/2012 10:54:14 PM | | max disk usage: 20.00GB
9/15/2012 10:54:14 PM | | max CPUs used: 2
9/15/2012 10:54:14 PM | | (to change preferences, visit the web site of an attached project, or select Preferences in the Manager)
9/15/2012 10:54:14 PM | | Not using a proxy
9/15/2012 10:54:15 PM | | Running CPU benchmarks
9/15/2012 10:54:15 PM | | Suspending computation - CPU benchmarks in progress
9/15/2012 10:54:46 PM | | Benchmark results:
9/15/2012 10:54:46 PM | | Number of CPUs: 2
9/15/2012 10:54:46 PM | | 2567 floating point MIPS (Whetstone) per CPU
9/15/2012 10:54:46 PM | | 8631 integer MIPS (Dhrystone) per CPU

Hope you can help me get GPU tasks from Moo! Wrapper running again. They did run under V6.12 before I started upgrading BOINC.

---

David - KC2WUF

Bump. I also have Multiple Nvidia cards on one of my machines. 3-9800GTX+. I can get Wu's but they never start. On my other machine I have 3 Ati cores in 2 cards. the App info xml fixed that problem. I have looked all over this forum without success...


How can I run multiple Nvidia cards with Moo?

Bump. I also have Multiple Nvidia cards on one of my machines. 3-9800GTX+. I can get Wu's but they never start. On my other machine I have 3 Ati cores in 2 cards. the App info xml fixed that problem. I have looked all over this forum without success...


How can I run multiple Nvidia cards with Moo?

Do you have a cc_config/xml file telling it to use all gpu's? If not put this in the C:\ProgramData\Boinc directory:
<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
</options>
</cc_config>

Copy and paste the above into NOTEPAD, NOT a word processing program!! Then save it as a txt type file as cc_config.xml. Basically it just tells Boinc to use all the gpu's it finds. The app_info files tell Boinc HOW to use the gpu's, such as run 2 units per gpu etc. You will have to restart Boinc and MAY lose all your existing units when you do, it is BEST to finish all your existing units PRIOR to doing this. ONE of my pc's did NOT trash all the existing units but that could be because I have mismatched gpu's in it.

Oh I almost forgot the 1 in the file means YES do that, while a zero would men NO do not do that. here is a link to other options for the cc_config.xml file:
https://boinc.berkeley.edu/w/?title=Client_configuration&amp;diff=3292&amp;oldid=prev

Unfortunately, I have same problem and it is not problem with

<use_all_gpus>1</use_all_gpus>

configuration.

Anyone have other suggestions?

Unfortunately, I have same problem and it is not problem with

<use_all_gpus>1</use_all_gpus>

configuration.

Anyone have other suggestions?

Do you have a Dummy Plug, or 2nd monitor plugged into the 2nd gpu? You can make your own Dummy Plug in 30 seconds using these directions:
http://www.overclock.net/t/384733/the-30-second-dummy-plug

You only need the 2nd monitor plugged in until the pc starts up then you can unplug it, Windows turns it off if there is nothing using it. The Dummy Plug gives it a little resistance making Windows think it is in use.

The last thing it could be is the lack of support for this at Moo, Teemu has said it is on his list to fix. After the recent crashes my dual gpu pc's did not work either and they are now at Collatz. It works for some people, just not everyone! And as I said it USED to work for me here, it just doesn't right now.

Do you have a Dummy Plug, or 2nd monitor plugged into the 2nd gpu? You can make your own Dummy Plug in 30 seconds using these directions:
http://www.overclock.net/t/384733/the-30-second-dummy-plug

I am not sure, that here is Dummy Plug problem, because other GPU projects, for example PrimeGrid or Collatz using both my GPUs without any problems. But maybe I'll try...

Hello,

Good news is that I managed to reproduce this problem on my dual nVidia system. Bad news is that the BOINC Client scheduling bug is still there on v7.0.42 but I managed to work around by using simple file described at http://boinc.berkeley.edu/trac/wiki/ClientAppConfig. It's much simpler to setup than a full anonymous platform. Here's the app_config.xml file I added to the ./projects/moowrap.net/ directory:



dnetc
1

1.00
0.05




You will need at least v7.0.40 to use that and client restart is required to activate it (client log will tell you when a valid file is found). Also be careful since this will allow other GPU projects run at the same time and you might end up overloading your graphic cards.

I will continue looking in to this problem but in the end, it's a BOINC Client scheduling bug that limits projects to max. 1 GPU resource per task.

-w

Thank You, Teemu, Moo now running on one of my GPUs.

By the way, ~30000 credits per day for GTX 470 looks like very small amount comparing with others GPU projects. I am not credit hunter, but, for example, DistrRTgen paying around 10x more.

Thank You, Teemu, Moo now running on one of my GPUs.

By the way, ~30000 credits per day for GTX 470 looks like very small amount comparing with others GPU projects. I am not credit hunter, but, for example, DistrRTgen paying around 10x more.

ATI gpu's working here better than Nvidia gpu's*.

*working better = give more points per hour.

Hello,

Good news is that I managed to reproduce this problem on my dual nVidia system. Bad news is that the BOINC Client scheduling bug is still there on v7.0.42 but I managed to work around by using simple file described at http://boinc.berkeley.edu/trac/wiki/ClientAppConfig. It's much simpler to setup than a full anonymous platform. Here's the app_config.xml file I added to the ./projects/moowrap.net/ directory:

<app_config>
<app>
<name>dnetc</name>
<max_concurrent>1</max_concurrent>
<gpu_versions>
<gpu_usage>1.00</gpu_usage>
<cpu_usage>0.05</cpu_usage>
</gpu_versions>
</app>
</app_config>

You will need at least v7.0.40 to use that and client restart is required to activate it (client log will tell you when a valid file is found). Also be careful since this will allow other GPU projects run at the same time and you might end up overloading your graphic cards.

I will continue looking in to this problem but in the end, it's a BOINC Client scheduling bug that limits projects to max. 1 GPU resource per task.

-w

Okay let me start this off by saying I have two AMD gpu's NOT Nvidia gpu's in the machine I about to talk about! I used the above app_config,xml file MINUS the line "<max_concurrent>1</max_concurrent>" to run TWO units at the same time in the machine. I made one other change to this line "<gpu_usage>1.00</gpu_usage>", I changed the 1.00 to 0.50

You will have a significantly better result using this:

<app_config>
<app>
<name>dnetc</name>
<gpu_versions>
<gpu_usage>0.50</gpu_usage>
<cpu_usage>0.15</cpu_usage>
</gpu_versions>
</app>
</app_config>

This is for GPUs with 1 Gb or more of memory. Note the higher value for cpu_usage; not changing the gpu_usage this alone will drop mean computing time down with a significant amount. Adding the gpu_usage value makes better use of the GPU and will drop the mean time even further. If you have two GPUs like Mickey this app_config.xml will let you run 4 WUs at a time.

---