So I cant use opti app to Run CPU? I was hoping to run cpu and gpu, but cpu is too slow without opti app.If I rename file will it still run opti app?

Let me think on it and test it...

Ok I'll check back in awhile. Thanks for your help Logan.
;-p

Sorry... I tryed some things but don't works...

Now I have other things 'to do'... Sorry if I can't help you now (by the moment).


Best regards.

---

Logan.

BOINC FAQ Service (Ahora, también disponible en Español/Now available in Spanish)

thx Logan, could you pls tell where exactly to look for that info?

thx Logan, could you pls tell where exactly to look for that info?

Of course...

In the project web page, select 'your account'. Next you can select 'view tasks'. when you find a 'completed task' make a click in the number called 'Task ID'. Then you can see the log of your task.

I.e. your task log

---

Logan.

BOINC FAQ Service (Ahora, también disponible en Español/Now available in Spanish)

So I cant use opti app to Run CPU? I was hoping to run cpu and gpu, but cpu is too slow without opti app.If I rename file will it still run opti app?

Let me think on it and test it...

Ok I'll check back in awhile. Thanks for your help Logan.
;-p

Sorry... I tryed some things but don't works...

Now I have other things 'to do'... Sorry if I can't help you now (by the moment).


Best regards.

Thanks for your time and effort Logan. Oh well it was worth a try anyways.
;-p

---

So I cant use opti app to Run CPU? I was hoping to run cpu and gpu, but cpu is too slow without opti app.If I rename file will it still run opti app?

Let me think on it and test it...

Ok I'll check back in awhile. Thanks for your help Logan.
;-p

Sorry... I tryed some things but don't works...

Now I have other things 'to do'... Sorry if I can't help you now (by the moment).


Best regards.

Thanks for your time and effort Logan. Oh well it was worth a try anyways.
;-p

I don't have any time for that ATM (sorry), but sure our friends can search a solution for that... Don't dessist...

---

Logan.

BOINC FAQ Service (Ahora, también disponible en Español/Now available in Spanish)

@Logan, here's the details

<core_client_version>6.6.36</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
Device index specified on the command line was 0
Looking for a Double Precision capable nVidia GPU
The device Device Emulation (CPU) from the command line cannot be used because a device supporting compute capability 1.3 (Double Precision) is required
Found 1 CUDA cards
Found a Device Emulation (CPU)
Device cannot be used, it does not have compute capability 1.3 support
No compute capability 1.3 cards have been found, exiting...

</stderr_txt>
]]>

I have NVidia GTX 280 but my drivers a bit old so I've updated to 190.62, now running a milkyway@home project on my GPU just freezes my machine right away (have tried it about 5-6 times) so I've disabled computation on GPU :'(

thx for your help, I'll give it another try with nVidia updates&patches the crappy 190.62 release (have a few other non-BOINC issues)

@Logan, here's the details

<core_client_version>6.6.36</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
Device index specified on the command line was 0
Looking for a Double Precision capable nVidia GPU
The device Device Emulation (CPU) from the command line cannot be used because a device supporting compute capability 1.3 (Double Precision) is required
Found 1 CUDA cards
Found a Device Emulation (CPU)
Device cannot be used, it does not have compute capability 1.3 support
No compute capability 1.3 cards have been found, exiting...

</stderr_txt>
]]>

I have NVidia GTX 280 but my drivers a bit old so I've updated to 190.62, now running a milkyway@home project on my GPU just freezes my machine right away (have tried it about 5-6 times) so I've disabled computation on GPU :'(

thx for your help, I'll give it another try with nVidia updates&patches the crappy 190.62 release (have a few other non-BOINC issues)

Isn't your computer fault. CUDA MW app now takes a lot of your user interface time... Are them who must to fix it...

Best regards.

---

Logan.

BOINC FAQ Service (Ahora, también disponible en Español/Now available in Spanish)

Thanks but my online project and personal settings were already set that way.
Still the same probs.

Curiously:
9/7/2009 5:07:23 PM Milkyway@home Sending scheduler request: To fetch work.
9/7/2009 5:07:23 PM Milkyway@home Requesting new tasks for GPU
9/7/2009 5:07:29 PM Milkyway@home Scheduler request completed: got 0 new tasks
9/7/2009 5:07:29 PM Milkyway@home Message from server: No work sent
9/7/2009 5:07:29 PM Milkyway@home Message from server: Can't use NVIDIA GPU app for MilkyWay@Home: Your GPU lacks the needed features


It's doing the exact opposite. It's requesting GPU WU's and not requesting CPU WU's.

I'll wait for a little more help before suspending this project and waiting for updates to it. Laudable & fascinating goals but I object to subjecting unwilling volunteers to bugs.

Thanks but my online project and personal settings were already set that way.
Still the same probs.

Curiously:
9/7/2009 5:07:23 PM Milkyway@home Sending scheduler request: To fetch work.
9/7/2009 5:07:23 PM Milkyway@home Requesting new tasks for GPU
9/7/2009 5:07:29 PM Milkyway@home Scheduler request completed: got 0 new tasks
9/7/2009 5:07:29 PM Milkyway@home Message from server: No work sent
9/7/2009 5:07:29 PM Milkyway@home Message from server: Can't use NVIDIA GPU app for MilkyWay@Home: Your GPU lacks the needed features


It's doing the exact opposite. It's requesting GPU WU's and not requesting CPU WU's.

I'll wait for a little more help before suspending this project and waiting for updates to it. Laudable & fascinating goals but I object to subjecting unwilling volunteers to bugs.

Well...

Check only the GPU work in 'your preferences'. And reset the project. Be sure you have a GTX2XX card and 190.38 driver or greater.

---

Logan.

BOINC FAQ Service (Ahora, también disponible en Español/Now available in Spanish)

I DETACH'd and then reattached the proj and things seem to be working but am keeping a leery eye on things.

@ Logan
I renamed the file and have a full cache of GPU WUs now, maybe if we find a way to make it work I just have to put the old name back.Anyway thanks again for your time and trouble.
;-p

---

What is the button on boincmanager > advance > preferences > use GPU while computer is in use?

If not clicked in computer waits until mv cuda start up to 10 min, and if clicked in, it start ok, based on preferens: 'In use' means mouse/keyboard activity in last 2 minutes or what ever...bla bla blaa..

That button don't make any different in other cuda projects, but here it do?
Why?

My GTX295 is doing the WUs in about 7 minutes under XP x64 sp2 w/Boinc 6.10.1 and 190.62(WHQL).

Thats strange.. I assembled my new computer today.. and run CUDA WU for the first time.. With low CPU usage compared with high CPU usage i did not notice much difference.. maybe 30-60sec on one GPU WU..

But my WU (the first 3-4 ones) finnied in 4min area.. And the rest 25 took around 5 minutes...

Thats strange? You have a GTX295? ...

My specs:
Core i7 920 @ Stock speed, GTX275, 6GB RAM .. Nothing special...
My GPU temp is around 75c when running in a room 25-26c ... And CPU temp with 100% CPU load and 8 tasks running is 55-59c in the same room.

EDIT: forgot to mention Boinc and OS
Boinc 6.6.36 OS: Windows7 64bit

@scibuff:

The device Device Emulation (CPU) from the command line cannot be used

Your CPU is using the emulation library for CUDA, hence boinc thinks it is a GPU, but the library has 'older' compute capability, so it is rejected.

I am not familiar with the command line, but some searching for 'emulation' might help. I recommend switching that option off, wherever it is set.

---

Team Bletchley Park.

So I cant use opti app to Run CPU? I was hoping to run cpu and gpu, but cpu is too slow without opti app.If I rename file will it still run opti app?

If I understand you correctly you want to run the optimized CPU app and the CUDA app in parallel.

Here is an app_info.xml which should just do that (adapted from lunatics). You have to change the name of the CPU app to the one which you are actually using (in my case this is the 64bit SSE3 version)! Stop Boinc, create an empty text file in your milkyway project directory (this is were also the optimized app is located), copy everything from below in it, rename the file to "app_info.xml", start Boinc (previously downloaded CUDA workunits will be discarded). Absolutely no guarantee that it will work - it works for me though...

<app_info>
 <app>
  <name>milkyway</name>
 </app>
 <file_info>
  <name>astronomy_0.20_x64_SSE3.exe</name>
  <executable/>
 </file_info>

 <app_version>
  <app_name>milkyway</app_name>
  <version_num>19</version_num>
  <file_ref>
   <file_name>astronomy_0.20_x64_SSE3.exe</file_name>
   <main_program/>
  </file_ref>
 </app_version>

 <app_version>
  <app_name>milkyway</app_name>
  <version_num>20</version_num>
  <file_ref>
   <file_name>astronomy_0.20_x64_SSE3.exe</file_name>
   <main_program/>
  </file_ref>
 </app_version>

 <app>
    <name>milkyway</name>
    </app>
    <file_info>
        <name>milkyway_0.20_windows_intelx86__cuda23.exe</name>
        <executable/>
    </file_info>
    <file_info>
        <name>cudart.dll</name>
        <executable/>
    </file_info>
    <file_info>
        <name>cutil32.dll</name>
        <executable/>
    </file_info>
    <app_version>
        <app_name>milkyway</app_name>
        <version_num>20</version_num>
        <plan_class>cuda23</plan_class>
        <avg_ncpus>0.040000</avg_ncpus>
        <max_ncpus>0.040000</max_ncpus>
        <coproc>
            <type>CUDA</type>
            <count>1</count>
        </coproc>
        <file_ref>
            <file_name>milkyway_0.20_windows_intelx86__cuda23.exe</file_name>
            <main_program/>
         </file_ref>
        <file_ref>
	       <file_name>cudart.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>cutil32.dll</file_name>
        </file_ref>        
    </app_version>

</app_info>

But works on 260 gpu's that don't crunch gpugid anymore, happy!

Yea, great, more Windoze only crap. Anybody wanna buy a 2 month old XFX GTX-260.

But works on 260 gpu's that don't crunch gpugid anymore, happy!

Yea, great, more Windoze only crap. Anybody wanna buy a 2 month old XFX GTX-260.

There is an issue associated with executing CUDA applications through BOINC when using Linux/GNU. The application will execute correctly when run through a terminal, however, when BOINC fork+execs the application it will segfault within NVIDIA's driver [1].

[1] http://forums.nvidia.com/index.php?showtopic=105178&mode=linearplus

But works on 260 gpu's that don't crunch gpugid anymore, happy!

Yea, great, more Windoze only crap. Anybody wanna buy a 2 month old XFX GTX-260.

There is an issue associated with executing CUDA applications through BOINC when using Linux/GNU. The application will execute correctly when run through a terminal, however, when BOINC fork+execs the application it will segfault within NVIDIA's driver [1].

[1] http://forums.nvidia.com/index.php?showtopic=105178&mode=linearplus

Since i'm experiencing the same trouble over at collatz, having boinc trash my cuda app, we now figured out that this issue only applies to the 6.6.x boinc clients.

The new 6.10.x clients work just fine and don't crash the cuda app any more.

---

Join Support science! Joinc Team BOINC United now!

Is there anything I can do to crunch CUDA while using my PC without having my screen blink everytime a WU is completed and the next one started?

I'm running 2x GTX260 under W7-64 Ultimate with the 6.6.38 client.

I'm running MW on Windows 7 32bit, BOINC 6.10.4, nVidia GTX275. There doesn't seem to be an optimized app for Cuda and therefore doesn't have an app_info so that I can set the "<count>.50</count>".

Is there something I need to do to get the most out of this box? Any help would be appreciated.

Thanks,

Jeff

---