Einstein@home News Q1 2011

BaBa · 7 Feb 2011 at 13:50

8 MILLION down

With the help of the new gpu app and app_info files im upto world No.32 for rac and will drop about 95k today

ToxicTBag · 7 Feb 2011 at 15:53

halz said:
OK, I have done the newsletter, made the sigs/wallpaper and I have just emailed approximately 270 members.
All the emails were directed to this post so come on peeps, it wouldnt take many more members to raise our output to the highest in the UK

WE NEED YOU

Okellydokellydo

Well it looks like theres been a few changes since i last attempted Einstein.
Have attached and am running 7 cpu tasks and 4 of the new BRP3 units.

Question time :rolleyes:

I was born to fiddle with stuff so where might one snatch an app_info from and what are peoples experiences with it, as in just how many units can i run at a time 2 maybe 3 ?

Any pointers or help would be much appreciated as allways

loudbob · 7 Feb 2011 at 15:57

ToxicTBag said:
Okellydokellydo

Well it looks like theres been a few changes since i last attempted Einstein.
Have attached and am running 7 cpu tasks and 4 of the new BRP3 units.

Question time I was born to fiddle with stuff so where might one snatch an app_info from and what are peoples experiences with it, as in just how many units can i run at a time 2 maybe 3 ?

Any pointers or help would be much appreciated as allways

Should be able to find one in the Einstein forums - that's where I got one.

Looks fairly long (but it works).

I'm running 3 wu's on a 430 - works fine and should net around 8k per day.

ToxicTBag · 7 Feb 2011 at 15:58

Ta muchly Mr Loudbob sah, and dont think for one minute that lets you off the photo hook

**Edit** It's all coming back now, my memory aint as bad as i thought. Could not find any app_info so just changed it in client_state voila

loudbob · 7 Feb 2011 at 15:59

Try this:

<app_info>
<app>
<name>einstein_S5GC1HF</name>
<user_friendly_name>Global Correlations S5 HF search #1</user_friendly_name>
</app>
<app>
<name>einsteinbinary_BRP3</name>
<user_friendly_name>Binary Radio Pulsar Search</user_friendly_name>
</app>
<file_info>
<name>einstein_S5GC1HF_3.06_windows_intelx86__S5GCESSE2.exe</name>
<main_program/>
</file_info>
<file_info>
<name>einstein_S5R6_3.01_graphics_windows_intelx86.exe</name>
<executable/>
</file_info>
<file_info>
<name>cufft32_23.dll</name>
<executable/>
</file_info>
<file_info>
<name>cudart32_23.dll</name>
<executable/>
</file_info>
<file_info>
<name>einsteinbinary_BRP3_1.04_windows_intelx86__BRP3cuda32.exe</name>
<executable/>
</file_info>
<file_info>
<name>einsteinbinary_BRP3_1.05_windows_intelx86__BRP3cuda32.exe</name>
<executable/>
</file_info>
<file_info>
<name>einsteinbinary_BRP3_1.00_graphics_windows_intelx86.exe</name>
<executable/>
</file_info>
<file_info>
<name>cudart_xp32_32_16.dll</name>
<executable/>
</file_info>
<file_info>
<name>cufft_xp32_32_16.dll</name>
<executable/>
</file_info>
<file_info>
<name>db.dev.win.4330b3e5</name>
</file_info>
<file_info>
<name>dbhs.dev.win.4330b3e5</name>
</file_info>
<file_info>
<name>db.dev.win.96b133b1</name>
</file_info>
<file_info>
<name>dbhs.dev.win.96b133b1</name>
</file_info>
<app_version>
<app_name>einsteinbinary_BRP3</app_name>
<version_num>104</version_num>
<platform>windows_intelx86</platform>
<avg_ncpus>0.200000</avg_ncpus>
<max_ncpus>1.000000</max_ncpus>
<plan_class>BRP3cuda32</plan_class>
<api_version>6.13.0</api_version>
<file_ref>
<file_name>einsteinbinary_BRP3_1.04_windows_intelx86__BRP3cuda32.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>cudart_xp32_32_16.dll</file_name>
<open_name>cudart32_32_16.dll</open_name>
<copy_file/>
</file_ref>
<file_ref>
<file_name>cufft_xp32_32_16.dll</file_name>
<open_name>cufft32_32_16.dll</open_name>
<copy_file/>
</file_ref>
<file_ref>
<file_name>einsteinbinary_BRP3_1.00_graphics_windows_intelx86.exe</file_name>
<open_name>graphics_app</open_name>
</file_ref>
<file_ref>
<file_name>db.dev.win.4330b3e5</file_name>
<open_name>db.dev</open_name>
<copy_file/>
</file_ref>
<file_ref>
<file_name>dbhs.dev.win.4330b3e5</file_name>
<open_name>dbhs.dev</open_name>
<copy_file/>
</file_ref>
<coproc>
<type>CUDA</type>
<count>0.330000</count>
</coproc>
<gpu_ram>220200960.000000</gpu_ram>
</app_version>
<app_version>
<app_name>einsteinbinary_BRP3</app_name>
<version_num>105</version_num>
<platform>windows_intelx86</platform>
<avg_ncpus>0.200000</avg_ncpus>
<max_ncpus>1.000000</max_ncpus>
<plan_class>BRP3cuda32</plan_class>
<api_version>6.13.0</api_version>
<file_ref>
<file_name>einsteinbinary_BRP3_1.05_windows_intelx86__BRP3cuda32.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>cudart_xp32_32_16.dll</file_name>
<open_name>cudart32_32_16.dll</open_name>
<copy_file/>
</file_ref>
<file_ref>
<file_name>cufft_xp32_32_16.dll</file_name>
<open_name>cufft32_32_16.dll</open_name>
<copy_file/>
</file_ref>
<file_ref>
<file_name>einsteinbinary_BRP3_1.00_graphics_windows_intelx86.exe</file_name>
<open_name>graphics_app</open_name>
</file_ref>
<file_ref>
<file_name>db.dev.win.96b133b1</file_name>
<open_name>db.dev</open_name>
<copy_file/>
</file_ref>
<file_ref>
<file_name>dbhs.dev.win.96b133b1</file_name>
<open_name>dbhs.dev</open_name>
<copy_file/>
</file_ref>
<coproc>
<type>CUDA</type>
<count>0.330000</count>
</coproc>
<gpu_ram>220200960.000000</gpu_ram>
</app_version>
<app_version>
<app_name>einstein_S5GC1HF</app_name>
<version_num>306</version_num>
<platform>windows_intelx86</platform>
<plan_class>S5GCESSE2</plan_class>
<api_version>6.13.0</api_version>
<file_ref>
<file_name>einstein_S5GC1HF_3.06_windows_intelx86__S5GCESSE2.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>einstein_S5R6_3.01_graphics_windows_intelx86.exe</file_name>
<open_name>graphics_app</open_name>
</file_ref>
</app_version>
</app_info>

Most likely, your GPU RAM won't match this App_Info - but neither does mine - and I have had no issues - all WU's have finished and validated with the exception of one (but I think it was an issue with validating against a CPU wu).

p.s. this uses cpu & gpu - so if you only want gpu wu's make sure you specify in your Einstein Account Options.

loudbob · 7 Feb 2011 at 16:05

ToxicTBag said:
Ta muchly Mr Loudbob sah, and dont think for one minute that lets you off the photo hook

Ahhh, yes.....

After much discussion with another etailer - I have finally managed to get a refund on my previous order.

Subsequent orders have now been placed (today) - with places which do have stock - should be expecting various deliveries over the next few days.

So expect the wc build to start at the weekend (latest).

BaBa · 7 Feb 2011 at 16:43

Im running 3 wu on a 1gb card and have had no problems so far but some people have picked up wu's that need 430mb each so could be a problem later.

ToxicTBag · 7 Feb 2011 at 19:51

When running 1 wu per gpu completion time is around 1 hour, i edited the client state to run 2 tasks but it's literally taking twice as long.
Mr Loudbobs app info above, should i save it as app_info.xml and just drop it into the Einstein folder...i is a bit puzzled

TheTross · 7 Feb 2011 at 19:59

Yep. Save it as app_info.xml, stop BOINC, drop the app_info file into the Einstein folder and restart BOINC. Just make sure you have all the files that are mentioned in app_info before restarting, as it may cause problems otherwise. (If you do need to download any missing files, go here and scroll down until you find the appropriate one(s), then stick them in the Einstein folder.)

By the way, if you don't use the graphics feature within BOINC then you can remove all lines that mention the graphics apps. It won't speed anything up but it'll make your app_info file a little less cluttered.

edgedemon · 7 Feb 2011 at 21:51

Nice news Mr Halz, might have to give this new CUDA app a go

ToxicTBag · 7 Feb 2011 at 22:55

Tried and failed sacking it for today.

Biffa · 7 Feb 2011 at 23:15

Wimp..

Toxic, be careful with that app_info.xml there are a lot of broken lines in it with spaces where they shouldn't be.

Biffa · 7 Feb 2011 at 23:36

OK I joined, probably going to ruin my Folding performance, but lets see.

OK I made a new app_info.xml which is simpler, this is ONLY FOR GPU, it doesn't do any CPU. But it does handle 1.04 and 1.05 WU's with the one binary.

Its set to do 2 WU's per GPU, if you want to do three (you need 1024MB of VRAM) then change <count>0.500000</count> in the following section:

PHP:

       <coproc>
            <type>CUDA</type>
            <count>0.500000</count>
        </coproc>

1.0 for 1, 0.5 for 2, 0.33 for 3, 0.25 for 4.

Its referenced twice in the file so make sure you change both.

PHP:

<app_info>
    <app>
        <name>einsteinbinary_BRP3</name>
        <user_friendly_name>Binary Radio Pulsar Search</user_friendly_name>
    </app>
    <file_info>
        <name>einsteinbinary_BRP3_1.05_windows_intelx86__BRP3cuda32.exe</name>
        <executable/>
    </file_info>
    <file_info>
        <name>cudart_xp32_32_16.dll</name>
        <executable/>
    </file_info>
    <file_info>
        <name>cufft_xp32_32_16.dll</name>
        <executable/>
    </file_info>
    <file_info>
        <name>db.dev.win.96b133b1</name>
    </file_info>
    <file_info>
        <name>dbhs.dev.win.96b133b1</name>
    </file_info>
    <app_version>
        <app_name>einsteinbinary_BRP3</app_name>
        <version_num>105</version_num>
        <platform>windows_intelx86</platform>
        <avg_ncpus>0.200000</avg_ncpus>
        <max_ncpus>0.200000</max_ncpus>
        <plan_class>BRP3cuda32</plan_class>
        <api_version>6.13.0</api_version>
        <file_ref>
            <file_name>einsteinbinary_BRP3_1.05_windows_intelx86__BRP3cuda32.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>cudart_xp32_32_16.dll</file_name>
            <open_name>cudart32_32_16.dll</open_name>
            <copy_file/>
        </file_ref>
        <file_ref>
            <file_name>cufft_xp32_32_16.dll</file_name>
            <open_name>cufft32_32_16.dll</open_name>
            <copy_file/>
        </file_ref>
        <file_ref>
            <file_name>db.dev.win.96b133b1</file_name>
            <open_name>db.dev</open_name>
            <copy_file/>
        </file_ref>
        <file_ref>
            <file_name>dbhs.dev.win.96b133b1</file_name>
            <open_name>dbhs.dev</open_name>
            <copy_file/>
        </file_ref>
        <coproc>
            <type>CUDA</type>
            <count>0.500000</count>
        </coproc>
            <gpu_ram>220200960.000000</gpu_ram>
    </app_version>
    <app_version>
        <app_name>einsteinbinary_BRP3</app_name>
        <version_num>104</version_num>
        <platform>windows_intelx86</platform>
        <avg_ncpus>0.200000</avg_ncpus>
        <max_ncpus>0.200000</max_ncpus>
        <plan_class>BRP3cuda32</plan_class>
        <api_version>6.13.0</api_version>
       <file_ref>
            <file_name>einsteinbinary_BRP3_1.05_windows_intelx86__BRP3cuda32.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>cudart_xp32_32_16.dll</file_name>
            <open_name>cudart32_32_16.dll</open_name>
            <copy_file/>
        </file_ref>
        <file_ref>
            <file_name>cufft_xp32_32_16.dll</file_name>
            <open_name>cufft32_32_16.dll</open_name>
            <copy_file/>
        </file_ref>
        <file_ref>
            <file_name>db.dev.win.96b133b1</file_name>
            <open_name>db.dev</open_name>
            <copy_file/>
        </file_ref>
        <file_ref>
            <file_name>dbhs.dev.win.96b133b1</file_name>
            <open_name>dbhs.dev</open_name>
            <copy_file/>
        </file_ref>
        <coproc>
            <type>CUDA</type>
            <count>0.500000</count>
        </coproc>
            <gpu_ram>220200960.000000</gpu_ram>
    </app_version>
</app_info>

ToxicTBag · 8 Feb 2011 at 00:50

Will give that a try in the morning Mr Biffa sah, been a long day and going to get some shut eye Cheers

derk · 8 Feb 2011 at 03:57

edgedemon said:
Nice news Mr Halz, might have to give this new CUDA app a go

Cheers, I have just got stuck into this myself, started with this

Code:

<coproc>
            <type>CUDA</type>
            <count>0.330000</count>

............and every gpu unit errored out after approx 6 to 7 minutes, so changed it to

Code:

<coproc>
            <type>CUDA</type>
            <count>0.500000</count>

............now they are are still ok after 30 minutes so fingers crossed they finish and validate.

Thanks for the tip Biffa

Update: they now complete but are invalid so going to knock it down again

loudbob · 8 Feb 2011 at 04:03

Tried my i7 930 on this - oops......

Temps hit +95 - so put it on hold until the new wc kit gets installed.

Moved back to CPDN to keep temps below 80.

Just running the i7 860 & GT430.

What a come down......2 x GTS250's, 3 x GTX460's & 2 x GTX470's to a GT430 & 8600GT - but need to get the wc'ing up'n'runnin before I consider new GPU's.

derk · 8 Feb 2011 at 05:57

Biffa said:
Its referenced twice in the file so make sure you change both.

Bugger, didnt see that first time round, oh well back we go :rolleyes:

Biffa · 8 Feb 2011 at 08:21

Cleaned up the app_info.xml again there were a few references to the old app in there that still needed clearing out, should have no errors when loading now. (The errors were cosmetic but still..)

halz: what gpus? I wouldn't run 0.5 on anything less thank a GTX460

loudbob · 8 Feb 2011 at 09:47

Biffa said:
Cleaned up the app_info.xml again there were a few references to the old app in there that still needed clearing out, should have no errors when loading now. (The errors were cosmetic but still..)

halz: what gpus? I wouldn't run 0.5 on anything less thank a GTX460

I was running 3 wu's on a GT430 - but for safety I have dropped it to 2.

Also, like you I have cleaned up the App_Info - removing all traces of old apps etc.

Running fine now - but showing the gpu wu's as an estimated 14hr completion time when it's actually 2.5hrs (bit annoying as majority of wu's are now cpu - and reached daily limit of 256 wu's).

Biffa · 8 Feb 2011 at 09:54

I'm going to crunch the 40 or so in my cache then stop. CPU usage is too high to use this at the same time as trying to do a bigadv on the cpu. Dropped PPD for the bigadv from 42K to 25K