SETI MB CUDA for Linux

[riofl]

all my uploads finished so i switched to 6.6.11. so far it is getting only cpu workunits but i suspect that is because i still have a ton of unfinished gpu workunits. it detects my hardware properly and it sends 1 wu to each device

LOL I just checked my servers page and although it does everything right locally it is just the opposite of all the others in reporting! instead of what others report 2-teslas, this one reports to the page or at least the page displays 2 gtx285! oh well. minor problem affecting nothing but my ego

will see what happens as cuda units get finished to see if it gets both.

so far it is looking really good!

and yes I am running app_info.xml using the AK cpu multibeam application and the cuda 2.2 vlar killer app. i also am running the cpu-gpu script but so far i do that manually until i get the ratios adjusted properly. set it too far before and wound up with no cpu apps and all gpu apps.

[riofl]

it got a few cuda workunits so it is fetching both. it appears 6.6.11 works fine.

[sunu]

LOL I just checked my servers page and although it does everything right locally it is just the opposite of all the others in reporting! instead of what others report 2-teslas, this one reports to the page or at least the page displays 2 gtx285! oh well. minor problem affecting nothing but my ego

In my case it reports correctly "[2] NVIDIA GeForce GTX 295 (895MB)"

it got a few cuda workunits so it is fetching both. it appears 6.6.11 works fine.

The big question is not if it gets cuda workunits, but if it ALSO gets workunits for cpu AK_V8.

[riofl]

yeah iti got near 100 cpu AK workunits before it got the  cuda ones. it is downloading both properly. it seems to work just fine.

[sunu]

yeah iti got near 100 cpu AK workunits before it got the  cuda ones. it is downloading both properly. it seems to work just fine.

Well, then, there is nothing more we could ask. 6.6.11 it is for multi-gpus in linux.

[Tye]

Well, then, there is nothing more we could ask. 6.6.11 it is for multi-gpus in linux.

I've been looking, but haven't found exactly how to write my app_info.xml to be able to crunch MB wus with both CPU and GPU - can you point me in the right direction?

Right now my app_info.xml lets me do the optimized AP with the CPU and the CUDA MB with the GPU only, but I do run other projects as well.

[sunu]

I also confirm that 6.6.11 gets both CPU and GPU multibeam workunits.

Tye, this is the seti section of my app_info.xml. Mind you that your cpu or your cuda client might have a different name:

    <app>
        <name>setiathome_enhanced</name>
    </app>
    <file_info>
        <name>AK_V8_linux64_ssse3</name>
        <executable/>
    </file_info>
    <file_info>
        <name>setiathome-6.08.CUDA_2.2_x86_64-pc-linux-gnu</name>
        <executable/>
    </file_info>
    <file_info>
        <name>libcudart.so.2</name>
        <executable/>
    </file_info>
    <file_info>
        <name>libcufft.so.2</name>
        <executable/>
    </file_info>
    <app_version>
        <app_name>setiathome_enhanced</app_name>
        <version_num>603</version_num>
        <flops>5634219710.66940475</flops>
        <avg_ncpus>1.0000</avg_ncpus>
        <max_ncpus>1.0000</max_ncpus>
    <file_ref>
        <file_name>AK_V8_linux64_ssse3</file_name>
        <main_program/>
    </file_ref>
    </app_version>
    <app_version>
        <app_name>setiathome_enhanced</app_name>
        <version_num>608</version_num>
        <plan_class>cuda</plan_class>
        <flops>19317324722.295102</flops>
        <avg_ncpus>1.0000</avg_ncpus>
        <max_ncpus>1.0000</max_ncpus>
    <coproc>
        <type>CUDA</type>
        <count>1</count>
    </coproc>
    <file_ref>
        <file_name>setiathome-6.08.CUDA_2.2_x86_64-pc-linux-gnu</file_name>
        <main_program/>
    </file_ref>
    <file_ref>
        <file_name>libcudart.so.2</file_name>
    </file_ref>
    <file_ref>
        <file_name>libcufft.so.2</file_name>
    </file_ref>
    </app_version>

[riofl]

is the flops section of the app info file really important? i'm not using it. my file is below. does the addition of the flops help with processing efficiency?

<app_info>
<app>
<name>setiathome_enhanced</name>
</app>
<file_info>
<name>AK_V8_linux64_ssse3</name>
<executable/>
</file_info>
<app_version>
<app_name>setiathome_enhanced</app_name>
<version_num>603</version_num>
<file_ref>
<file_name>AK_V8_linux64_ssse3</file_name>
<main_program/>
</file_ref>
</app_version>
<app>
<name>setiathome_enhanced</name>
</app>
<file_info>
<name>setiathome-6.08.CUDA_2.2_x86_64-pc-linux-gnu</name>
<executable/>
</file_info>
<file_info>
<name>libcudart.so.2</name>
<executable/>
</file_info>
<file_info>
<name>libcufft.so.2</name>
<executable/>
</file_info>
<app_version>
<app_name>setiathome_enhanced</app_name>
<version_num>608</version_num>
<plan_class>cuda</plan_class>
<avg_ncpus>0.250000</avg_ncpus>
<max_ncpus>0.250000</max_ncpus>
<coproc>
<type>CUDA</type>
<count>1</count>
</coproc>
<file_ref>
<file_name>setiathome-6.08.CUDA_2.2_x86_64-pc-linux-gnu</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>libcudart.so.2</file_name>
</file_ref>
<file_ref>
<file_name>libcufft.so.2</file_name>
</file_ref>
</app_version>
</app_info>

[riofl]

slightly off topic but possibly relevant.

does the cpu-gpu perl script V5 available in another topic actually catch vlars and vhars? maybe it doesn't report them because i have not seen it report any yet. a little concerned because i have had several computation error workunits from the 'vlar-killer 2.2 cuda' ap and with task viewing turned off at seti i cannot tell.

the other thing is would it be better to have my ratio set so that boinc never sees a shortage of cuda workunits so it never gets cuda workunits but only cpu workunits and then let this script supply the gpu work? i would think if it handles vlars it would be the most efficient way to never get one scheduled for cuda?

[sunu]

is the flops section of the app info file really important? i'm not using it. my file is below. does the addition of the flops help with processing efficiency?

It is important, but not for processing efficiency. flops numbers help boinc to better calculate estimated computation times, so to better plan ahead and download or not new workunits to fill the cache. Fairly accurate estimated computation times help stabilize Duration Correction Factor and brings boinc to a balanced state.

It is also the other way around. You see, Duration Correction Factor, estimated computation times, balanced boinc, good cache management are all interconnected and flops numbers in app_info.xml help to bring all these in a balanced state.

If you think that your boinc is balanced and steady without the flops numbers, then probably you don't need them. Otherwise you have to put them in.

Edit:

does the cpu-gpu perl script V5 available in another topic actually catch vlars and vhars? maybe it doesn't report them because i have not seen it report any yet. a little concerned because i have had several computation error workunits from the 'vlar-killer 2.2 cuda' ap and with task viewing turned off at seti i cannot tell.

I haven't used that script. Might be that vlar_kill has a greater vlar angle range than the script is looking for? To see if it is working correctly, after a script run that doesn't report anything, make a manual search in the workunits directory to see how many files(workunits) contain the text <true_angle_range>0.0 and then see if some of them are still assigned to cuda.

the other thing is would it be better to have my ratio set so that boinc never sees a shortage of cuda workunits so it never gets cuda workunits but only cpu workunits and then let this script supply the gpu work? i would think if it handles vlars it would be the most efficient way to never get one scheduled for cuda?

In both cases you don't avoid using the script. Choose whatever suits you better.

[riofl]

wow i've never given any of that any thought. never even knew of the term balanced boinc. i just install it let it do its thing. it seems to get and processand send workunits ok but isuppose there is something to be said for 'balancing it out'. maybe ill mess with that a bit.

time will tell on the script. ill search for that phrase in what i have now cause i dont like killing off units like that and my tesla cannot take them. its an early engineering version of it and it locks right up on vlars. this is the main reason i chose to run the vlar killer app to protect it.

[sunu]

my tesla cannot take them. its an early engineering version of it and it locks right up on vlars. this is the main reason i chose to run the vlar killer app to protect it.

Every GPU, in every OS, windows or linux has problems with VLARs.

[Tye]

I see this line:

<file_name>setiathome-6.08.CUDA_2.2_x86_64-pc-linux-gnu</file_name>

But I only have the old app without the 2.2 part - where are you getting this CUDA app?

[sunu]

I see this line:

<file_name>setiathome-6.08.CUDA_2.2_x86_64-pc-linux-gnu</file_name>

But I only have the old app without the 2.2 part - where are you getting this CUDA app?

Get it from http://calbe.dw70.de/mb/viewtopic.php?f=9&t=110

[riofl]

I see this line:

<file_name>setiathome-6.08.CUDA_2.2_x86_64-pc-linux-gnu</file_name>

But I only have the old app without the 2.2 part - where are you getting this CUDA app?

sunu reported the url, but a reminder you need to install the 2.2 cuda sdk and tools too and update the libs in every directory you have them from the sdk lib dir. especially /usr/lib64, the boinc dir and  the seti project dir and make sure those paths are in ld.so.conf and rerun ldconfig.