Laptop with 8400M-G

[bronan]

Hi guys i installed boinc on my little lappy.
Its a acer with TL-60 2 Ghz cpu and 4 Gb ddr2 nvida cuda installed driver 179.48
highest version available for this GPU it says to have 8 stream cpus at 400 mhz, 800 mhz shader and memory is at 800 Mhz. 256 Mb Dedicated and 768 Mb assigned shared memory available

Now when i run boinc lets say standard with seti@home non optimized it runs a cuda unit without problems and finish them withing reasonable time.
As second project i run MW@H but at the moment it just runs 1 each on 1 cpu so 1 S@H and 1 MW@H

Now i thought maybe i can speed things up and use the optimized apps posted here and finish S@H faster, here starts the problems non of the programs i tested runs the units without problems.
Every time the unit start the calculation and crashes after about 30 seconds ending in computation errors.
I resetted back to no opti and hoppa runs without errors again, do i need a special version or is it simply not possible.
The TL-60 supports SSE3 Amd optimized MW apllications without problems

[Claggy]

Hi guys i installed boinc on my little lappy.
Its a acer with TL-60 2 Ghz cpu and 4 Gb ddr2 nvida cuda installed driver 179.48
highest version available for this GPU it says to have 8 stream cpus at 400 mhz, 800 mhz shader and memory is at 800 Mhz. 256 Mb Dedicated and 768 Mb assigned shared memory available

Now when i run boinc lets say standard with seti@home non optimized it runs a cuda unit without problems and finish them withing reasonable time.
As second project i run MW@H but at the moment it just runs 1 each on 1 cpu so 1 S@H and 1 MW@H

Now i thought maybe i can speed things up and use the optimized apps posted here and finish S@H faster, here starts the problems non of the programs i tested runs the units without problems.
Every time the unit start the calculation and crashes after about 30 seconds ending in computation errors.
I resetted back to no opti and hoppa runs without errors again, do i need a special version or is it simply not possible.
The TL-60 supports SSE3 Amd optimized MW apllications without problems

You should try the Stock 6.08 Cuda app, since Raistmer's apps require drivers of v180.48 or later,
I also have a Laptop with a Nvidia GPU, It's a Samsung R700 with a 8400M GS, but only has 128Mb,
so won't run the app on the GPU, and goes into CPU fallback mode, very very slowly, even slower the stock 6.03,
Keep an eye out for that if that 256Mb ends up as being reported as 254Mb.

Claggy

Edit: after reading your post again, i realise you were running stock app before O.K, and are again, what Boinc version? and can you link to the host?

[bronan]

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4800886

This is the host running
I have tested a so called fixed (modded) 180.70 version nvidia driver but those crash the laptop.
Boinc versions tested 6.4.5,6.4.7,6.5.0 and 6.6.12
The 179.48 is the latest real version for laptops available.

[Claggy]

http://setiathome.berkeley.edu/show_host_detail.php?hostid=4800886

This is the host running
I have tested a so called fixed (modded) 180.70 version nvidia driver but those crash the laptop.
Boinc versions tested 6.4.5,6.4.7,6.5.0 and 6.6.12
The 179.48 is the latest real version for laptops available.

Have you got a cc_config.xml with this in it?

<cc_config>
<options>
<ncpus>3</ncpus>
</options>
</cc_config>

If you haven't, this will make it seem to Boinc that you have 3 cores,
so the two cores and the GPU are utilised at the same time,
but has a tendency to try and do 3 CPU tasks with Boinc 6.4.x,
put it in your BoincData folder,  (Note, it's not needed for Boinc 6.6.x)
Have you got the stock cuda app in your app_info?, do you need it?,
and do you need the links to the files?

Claggy

[bronan]

i just put the editted cc_config.xml on that machine and it runs fine now with 1 cuda stock app
but its ofcourse damn slow compared to the optimized ones.
Right after i started boinc fresh it started 2 cpu tasks and 1 cuda task so that works as it should also
As i can see it has started s@h enhanced 6.08 (cuda) as usual
and i am running 2 other projects on the cpu's, s@h gives way to low credits on cpu aps
or the time to complete them takes way too long for finish them in time.

[Claggy]

If you get the combined AK_V8 and Optimised Astropulse Package from here:

Win32_AK_AP_SSE3.7z

and these files:

setiathome_6.08_windows_intelx86__cuda.exe

libfftw3f-3-1-1a_upx.dll

cudart.dll

cufft.dll

Shut down Boinc, and drop the 4 Cuda files from above, and the two Astropulse apps from the package into your setiathome folder,

with this app_info.xml:

<app_info>
    <app>
        <name>setiathome_enhanced</name>
    </app>
    <file_info>
        <name>setiathome_6.08_windows_intelx86__cuda.exe</name>
        <executable/>
    </file_info>
    <file_info>
        <name>cudart.dll</name>
        <executable/>
    </file_info>
    <file_info>
        <name>cufft.dll</name>
        <executable/>
    </file_info>
    <file_info>
        <name>libfftw3f-3-1-1a_upx.dll</name>
        <executable/>
    </file_info>
    <app_version>
        <app_name>setiathome_enhanced</app_name>
        <version_num>603</version_num>
        <plan_class>cuda</plan_class>
        <api_version>6.3.22</api_version>
        <avg_ncpus>0.114729</avg_ncpus>
        <max_ncpus>0.114729</max_ncpus>
        <coproc>
            <type>CUDA</type>
            <count>1</count>
        </coproc>
        <file_ref>
            <file_name>setiathome_6.08_windows_intelx86__cuda.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>cudart.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>cufft.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>libfftw3f-3-1-1a_upx.dll</file_name>
        </file_ref>
    </app_version>
    <app_version>
        <app_name>setiathome_enhanced</app_name>
        <version_num>608</version_num>
        <plan_class>cuda</plan_class>
        <api_version>6.3.22</api_version>
        <avg_ncpus>0.114729</avg_ncpus>
        <max_ncpus>0.114729</max_ncpus>
        <coproc>
            <type>CUDA</type>
            <count>1</count>
        </coproc>
        <file_ref>
            <file_name>setiathome_6.08_windows_intelx86__cuda.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>cudart.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>cufft.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>libfftw3f-3-1-1a_upx.dll</file_name>
        </file_ref>
    </app_version>
    <app>
        <name>astropulse</name>
    </app>
    <file_info>
        <name>ap_5.00r103_SSE3.exe</name>
        <executable/>
    </file_info>
    <app_version>
        <app_name>astropulse</app_name>
        <version_num>500</version_num>
        <file_ref>
            <file_name>ap_5.00r103_SSE3.exe</file_name>
            <main_program/>
        </file_ref>
    </app_version>
    <app>
        <name>astropulse_v5</name>
    </app>
    <file_info>
        <name>ap_5.03r112_SSE3.exe</name>
        <executable/>
    </file_info>
    <app_version>
        <app_name>astropulse_v5</app_name>
        <version_num>503</version_num>
        <file_ref>
            <file_name>ap_5.03r112_SSE3.exe</file_name>
            <main_program/>
        </file_ref>
    </app_version>
</app_info>

You don't need any of the AK_V8 files, and the app_info has 603 entries, so any CPU Multibeam work will continue on the GPU.

Claggy

[bronan]

Thanx gonna fix this in the morning

[RottenMutt]

I am using 180+ drivers from http://www.laptopvideo2go.com/ on my HP laptop (dual core 2.0GHz with 9600M-512MB ddr2).  I'm running 6.6.17 client, Raistmer's V10 optimized package along with an edited MarkJ's app_info file with Raistmer's apps.  see http://setiathome.berkeley.edu/forum_thread.php?id=52589.
I've also oc'd my video card from 500MHz to 600MHz.  so far stable, temps are around 60C on gpu and 45C on cpu with a room temp around 70F.

[bronan]

Well i am gonna try both.
But to see a cuda app taking almost 6 hours to finish is not really rewarding , seti rewards on the fastest calculation.
Sadly there is no project which i can run which is usefull with this lappy.
The 8600 is much faster then my v.c. its only 399 mhz and can't be overclocked.
Second the h4cked nvidia drivers needs as stated a higher model number which supports cuda 1.3 and i think mine only supports 1.1 as far as i can find.
Still i can't complain about it with 500 GB disk space,4 Gb ram memory and able to play most games i think its still a good lappy
Just hoping to find the right optimized combo to get it to run a bit faster, on some projects.
On MW@home i can run the SSE3 optimized toys very well just with seti i seem not to get it running smooth.

[Claggy]

Nvidia has released Beta 185.81 drivers for Notebooks,
so anyone who couldn't run Raistmer's app's, now can run them.

Claggy