SETI MB CUDA for Linux

[ML1]

OK... To jump in with both feet and an AMD Athlon64 X2 + nVidia 8600 GT and...

Read all about the fun on s@h: Boinc 6.6.2 just released (CUDA)

Briefly, it appears to be working with two nice 19 tasks and additionally a nice 10 s@h CUDA task. Except that that the Boinc Manager consistently crashes whilst starting. The boinc client runs on unperturbed...

OK, first few results listed.

All pending or validated ok. No fails so far.

However, looks like out-of-memory for processing the gaussians.

What to check next?

Cheers,
Martin

[sunu]

Hmm, 256MB seem a bit borderline in your case. I've got a GTX 280 and I've never had out of memory errors. Do you have any compiz 3d effects enabled?

[ML1]

Hmm, 256MB seem a bit borderline in your case. I've got a GTX 280 and I've never had out of memory errors. Do you have any compiz 3d effects enabled?

Nope.

Simply 1280x1024, no effects, 4 desktops, running Mandriva 2009.0 with KDE4.

Any diagnostics that can be done?

Cheers,
Martin

[Crunch3r]

Hmm, 256MB seem a bit borderline in your case. I've got a GTX 280 and I've never had out of memory errors. Do you have any compiz 3d effects enabled?

Nope.

Simply 1280x1024, no effects, 4 desktops, running Mandriva 2009.0 with KDE4.

Any diagnostics that can be done?

Cheers,
Martin

I've seen that once. It happens when boinc suspends a WU to crunch another one in high priority mode. You need to make sure not to leave apps in memory while suspended.

When that happens again, you should check if there's another suspended instance of the cuda app present.

[smurf]

I'm using the latest official stable drivers 180.22. Go to http://www.nvidia.com/object/linux_display_ia32_180.22.html to grab them.
Have you overclocked your card? Do you know what temperatures it gets when running cuda?

Now I updated to 180.22. But same situation. The card is not overclocked and temperature is 62 C. There are two types of cases:
- small WUs with a time to completion of ~7 min are running fine without any errors and granted credit is the same as claimed credit (14-15)
- big WUs with a time to completion of ~25 min have the gausfit error message and the granted credit is smaller than the claimed credit (claimed ~50, granted ~40)

I also tried it with a good old TWM only desktop which should not block any resources of the GPU for desktop effects. But same situation.

I also tried seti_home in text console standalone mode, but then all WUs make lots of strange errors. Is CUDA only supported from within an X session ?

[koubi89]

sam 24 jan 2009 21:00:06 CET||Internet access OK - project servers may be temporarily down.
sam 24 jan 2009 21:01:08 CET||Project communication failed: attempting access to reference site
sam 24 jan 2009 21:01:08 CET|SETI@home|Temporarily failed upload of 16dc08ab.25701.18068.7.8.36_0_0: HTTP error
sam 24 jan 2009 21:01:08 CET|SETI@home|Backing off 1 min 0 sec on upload of 16dc08ab.25701.18068.7.8.36_0_0
sam 24 jan 2009 21:01:08 CET|SETI@home|Started upload of 16dc08ab.25701.18068.7.8.47_1_0
sam 24 jan 2009 21:01:09 CET||Internet access OK - project servers may be temporarily down.

an idea?

[sunu]

For the last day or so berkeley's servers are taking a hit and they are a bit slow. There is nothing on your end to worry about. More or less everyone has a problem uploading.

[arkayn]

The server is getting blasted, partly because of the new server bug.



When that green line goes back down everything will work better.

[koschi]

I also tried seti_home in text console standalone mode, but then all WUs make lots of strange errors. Is CUDA only supported from within an X session ?

CUDA (of couse) only works when the Nvidia driver is loaded. This is usually done during start of the X server/subsystem.  I guess the driver wasn't loaded/unloaded when you started/switched to standalone mode...

[Hefto99]

Hi everybody,

I'm not able to convince BOINC client 6.4.5 that my GK 8600 GT is CUDA capable (No CUDA devices found)
I have got: openSUSE 11.1 64-bit, latest nvidia driver 180.22, libraries are copied into /usr/lib64:

ldd setiathome-CUDA-6.08.x86_64-pc-linux-gnu
        linux-vdso.so.1 =>  (0x00007fff337ff000)
        libcufft.so.2 => /usr/lib64/libcufft.so.2 (0x00007f602b259000)
        libcudart.so.2 => /usr/lib64/libcudart.so.2 (0x00007f602b01b000)
        libcuda.so.1 => /usr/lib64/libcuda.so.1 (0x00007f602ab7b000)
        libstdc++.so.6 => /usr/lib64/libstdc++.so.6 (0x00007f602a86f000)
        libm.so.6 => /lib64/libm.so.6 (0x00007f602a619000)
        libpthread.so.0 => /lib64/libpthread.so.0 (0x00007f602a3fd000)
        libc.so.6 => /lib64/libc.so.6 (0x00007f602a0a4000)
        libdl.so.2 => /lib64/libdl.so.2 (0x00007f6029ea0000)
        libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007f6029c88000)
        librt.so.1 => /lib64/librt.so.1 (0x00007f6029a7f000)
        libz.so.1 => /lib64/libz.so.1 (0x00007f6029869000)
        /lib64/ld-linux-x86-64.so.2 (0x00007f602b573000)
ldd boinc
        linux-vdso.so.1 =>  (0x00007fffefdff000)
        libnsl.so.1 => /lib64/libnsl.so.1 (0x00007fc3e7857000)
        libdl.so.2 => /lib64/libdl.so.2 (0x00007fc3e7653000)
        libz.so.1 => /lib64/libz.so.1 (0x00007fc3e743d000)
        libstdc++.so.6 => /usr/lib64/libstdc++.so.6 (0x00007fc3e7131000)
        libpthread.so.0 => /lib64/libpthread.so.0 (0x00007fc3e6f15000)
        libm.so.6 => /lib64/libm.so.6 (0x00007fc3e6cbf000)
        libc.so.6 => /lib64/libc.so.6 (0x00007fc3e6966000)
        /lib64/ld-linux-x86-64.so.2 (0x00007fc3e7a6f000)
        libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007fc3e674e000)



Do you have any idea what else do I miss?

Thanks a lot for suport,
H99

[Crunch3r]

I recommend reading this post at the S@H MB http://setiathome.berkeley.edu/forum_thread.php?id=51556&nowrap=true#856583

Seems to me that you ran into the same issues.

HTH

[Hefto99]

Yeah, I have tried this before, but without sucecss 
I will try to reinstall the nvidia driver again if it helps...

[edit] I have reinstalled the driver and CUDA is detected now  Previous driver was installed from nvidia repository for SUSE, maybe this was the problem. [/edit]

H99

[Hefto99]

I have another question:
CPU utilization is still almost 100%  for SETI and temp. on the GK stays low (50 C).

Task Status shows " Running (0.04% CPUs, 1 CUDA)", progress proceeds very slowly.

Is it possible that CPU is too weak to feed Graphic Card (Athlon X2 3800+) ?

[sunu]

Do you have any MB work to run? Maybe you're running only AP?

[Hefto99]

Yes, app is setiathome_enhanced 6.08 (cuda).
I have set AP to NO in the seti prefs and using app_info.xml without AP.