Looks like system RAM, not GPU's.
Try both DLL sets and use what best suited to your system.
If this is new RAM check it with some memory test utility...

Are you sure? ;-D

..kidding..


The 800 MHz RAM ran @ 944 MHz @ 5-5-5-18 @ 2.1 V - in the QX6700 rig for 1 1/2 years. [with active cooling]
Now in the AMD 940 BE rig @ 800 MHz @ 5-5-5-18 @ 1.8 V* [all AUTO]
This are Corsair xms2 6400C4 sticks. 5-5-5-18 @ 1.8 V stock or 4-4-4-12 @ 2.1 V .
With memtest [@ QX6700 rig in past] I had some [~ 3 - 5] errors only @ 4-4-4-12, 5-5-5-18 worked fine. [800 MHz]
O.K., a well reason for warranty change in past.. but the Core2 was better/faster with 944/5-5-5-18 as 788/4-4-4-12, so it was O.K. ..

If I remember correct, since your 'GPU WU start opt.' I have this errors.

I'm the only one..? If yes - O.K., then it's my RAM.. ;-)

[* I guess the voltage]
Is there a prog with which I can see the RAM voltage in Windows?
Current at AUTO, so maybe not enough or to low voltage..


Since BOINC V6.6.20 your nice team mod isn't needed anymore.
Maybe you can open a new thread with the latest available CUDA versions? [V10/11]
[incl. dll's and needed app_info.xml]

Or maybe a hint to the app_info.xml thread?
app_info for AP500, AP503, MB603 and MB608


Which/where are the .dll-versions from you?
Please could you post where to find?
Or this are the .dll's from your V7 and V10 mod. This are the only different .dll-versions from you?
[in the 1st posts of the threads @ lunatics.kwsn.net?]

I could use the stock Berkeley .dll's also with your mod?
Or what's different?
I have no knowledge for what the .dll's are needed.. I'm not the profi.. I'm an 'amateur'.. ;-)

---

http://setiathome.berkeley.edu/result.php?resultid=1195495666
icfft=86040, PoT_activity=0, PoT_freq_bin=-1SETI@home error -12 Unknown error

http://setiathome.berkeley.edu/result.php?resultid=1190297258
icfft=94365, PoT_activity=0, PoT_freq_bin=-1SETI@home error -12 Unknown error

[This error I posted already in this thread]
http://setiathome.berkeley.edu/result.php?resultid=1195160891
icfft=86665, PoT_activity=0, PoT_freq_bin=-1SETI@home error -12 Unknown error

Now my 4th..
http://setiathome.berkeley.edu/result.php?resultid=1202303914
icfft=84509, PoT_activity=0, PoT_freq_bin=-1SETI@home error -12 Unknown error


[@ all - I'm the only one which see this errors? *confused*]

Or I'm the only one which see/look to the error column in the overview. ;-D [Now it's easy to find because of the choice of the look]
But need also some time, because of the 'bad WU header's '.. ;-)

---

[@ all - I'm the only one which see this errors? *confused*]

Or I'm the only one which see/look to the error column in the overview. ;-D [Now it's easy to find because of the choice of the look]
But need also some time, because of the 'bad WU header's '.. ;-)

You're not quite alone, I reported one here.

They certainly are easier to find with the task list filter. Also, I don't even look at ones with <1 sec CPU time anymore, since they all seem to be VLAR kills.

---

I've only seen Error -6 when I ran two gpus of My GTX295 video card and processed 2D video at the same time and yes this was at the 295's stock speed too, So I went back to using 1 gpu instead and have overclocked the GTX295 to 641MHz(briefly tried 648MHz but got an error -6 according to BV 1.4.2), Fan is on 100% and ram(1101MHz) and shaders(1382MHz) are overclocked and I'm seeing no problems.

For Raistmer it's for interest the -12 error.
Because it's a BUG in the current CUDA code. [stock - and because of this also in his app]

You let run the fan at 100 % ??
Woohoo.. I hope the rig is very far from the place you live..
Because I read the fan is veeery loud..

Because of this I bought 2 x GTX260 Core216 and not a GTX295 - the GPUs are much hotter and the fan must turn much faster for 2 GPUs as only 1 GPU in the same GPU case.

---

You're not quite alone, I reported one here.

They certainly are easier to find with the task list filter. Also, I don't even look at ones with <1 sec CPU time anymore, since they all seem to be VLAR kills.

Ops.. I didn't saw..

You are right.. the ~ 1 sec. errors are only from the 'VLAR kill'..
The -12 error occurs in the task process.. so you see some seconds of processing..

---

Oh contraire, It's not that loud, The ATi cooling fan is much louder and so My Delta 120mm 150cfm cpu fan is slightly louder than the 295's fan and yet less loud than the ATi fan, So It's not too bad as I can hear the hdd when the head is moving around inside the hdd.

Hmm.. it depend what you feel as loud.. ;-)

If your Delta with 150 cfm - this must be 255 m³/h [for the europeans/germans around here.. ;-)]..
..then I guess it must be ~ 45 dBA.. huuhh.. too loud for me.. ;-D

Everything > 25 dBA is toooo loud for me.. :-)

---

Since BOINC V6.6.20 your nice team mod isn't needed anymore.
Maybe you can open a new thread with the latest available CUDA versions? [V10/11]
[incl. dll's and needed app_info.xml]

Or maybe a hint to the app_info.xml thread?
app_info for AP500, AP503, MB603 and MB608


Which/where are the .dll-versions from you?
Please could you post where to find?
Or this are the .dll's from your V7 and V10 mod. This are the only different .dll-versions from you?
[in the 1st posts of the threads @ lunatics.kwsn.net?]

I could use the stock Berkeley .dll's also with your mod?
Or what's different?
I have no knowledge for what the .dll's are needed.. I'm not the profi.. I'm an 'amateur'.. ;-)

Yes, a new thread is really required for Raistmer's V10/V11 app in Boinc 6.6.20 standalone mode,
and a new download with the required .dll's so at least you get everything in one package. (If you really need them, or if you can make do if stock .dll's)

Claggy

Question
I seem to recall seeing a 64bit Cuda opp app but cant find the link,is there any advantage to running this app?
I have v10 running at the moment and all is well.
I have the app for vlar kill but I get errors when BM starts.
What am I doing wrong?
Sorry if this is the wrong place to post this.

---

Raistmer.. sorry, I hope you don't feel spammed from my posts.. [many in a row]
But maybe I found something interesting?



Hmm.. what happened here.. same MB AR WU but 618.8* (CPU 100.625) sec. and 301.8* (CPU 66.21875) sec.
[* Wall-clock time elapsed since last restart]

http://setiathome.berkeley.edu/result.php?resultid=1203882723

http://setiathome.berkeley.edu/result.php?resultid=1203882724



Also..
Normally a MB AR WU 0.44x - ~ 480 sec. - ~ 8 min. [~ 52 credits**]

0.547505 - 868.6 sec. - ~ 14.5 min. [45 credits**]
http://setiathome.berkeley.edu/result.php?resultid=1203924655

1.066794 - 1042.6 sec. - ~ 17.5 min. [30 credits**]
http://setiathome.berkeley.edu/result.php?resultid=1204001983

[** incl. CUDA overclaim!]

Is this normal?
AFAIK, < AR means > calculation time .
This is now exactly the contrary.. *confused*


Finally..
Or is the CPU/real wall clock time counter broken in BOINC V6.6.20 ?
I wasn't in front of the PC.. so I have only this counter times.


EDIT:
Or is this again my in past seen 1/2 CUDA performance lost with > BOINC V6.6.11 and your V7 and V10 CUDA app?
Current I use BOINC V6.6.20 .

---

Question
I seem to recall seeing a 64bit Cuda opp app but cant find the link,is there any advantage to running this app?
I have v10 running at the moment and all is well.
I have the app for vlar kill but I get errors when BM starts.
What am I doing wrong?
Sorry if this is the wrong place to post this.

Only CUDA 32 bit available.
On GPU you don't have 32 or 64 bit.

Or you mean the applications for the CPU?


If you have a bunch of VLARs, they will be 'killed' immediately..

Or all ARs will be errored?

---

I did x64 CUDA MB build but there was no speed increase so not worth to support it for now. Maybe later it will show better results.

About CUDA performance - if the core where CUDA MB executed is occupied with some higher priority process (with priority higher than CUDA MB worker thread one) the app can experience substantional performance degradation. That's why CPU affinity mod was created (to give user with multicore CPUs ability to bound other processes on higher core numbers while leaving first core mostly for CUDA MB.) Other BOINC apps executed on even more low priority and should not interfere with CUDA MB.

Thanks for the info.
Can you tell me politly where MB_6.08_mod_VLAR_kill_CUDA.exe is put in the app_info.xml?
Here is my info file.I have input the flops do I have to put the numbers after the decimal point in? Does it look ok to run?

<app_info>
<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>
<flops>5233741404.0081975</flops>
<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>
<flops>6047878955.742806</flops>
<file_ref>
<file_name>ap_5.03r112_SSE3.exe</file_name>
<main_program/>
</file_ref>
</app_version>
<app>
<name>setiathome_enhanced</name>
</app>
<file_info>
<name>AK_v8b_win_SSE3_AMD_GPU_CPU_team_V10.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>
<file_info>
<name>MB_6.08_mod_CUDA_V10.exe</name>
<executable/>
</file_info>

<app_version>
<app_name>setiathome_enhanced</app_name>
<version_num>528</version_num>
<file_ref>
<file_name>AK_v8b_win_SSE3_AMD_GPU_CPU_team_V10.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>
<file_ref>
<file_name>MB_6.08_mod_CUDA_V10.exe</file_name>
</file_ref>
</app_version>

<app_version>
<app_name>setiathome_enhanced</app_name>
<version_num>603</version_num>
<flops>4070687758.6730425</flops>
<file_ref>
<file_name>AK_v8b_win_SSE3_AMD_GPU_CPU_team_V10.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>
<file_ref>
<file_name>MB_6.08_mod_CUDA_V10.exe</file_name>
</file_ref>
</app_version>


<app_version>
<app_name>setiathome_enhanced</app_name>
<version_num>607</version_num>
<file_ref>
<file_name>AK_v8b_win_SSE3_AMD_GPU_CPU_team_V10.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>
<file_ref>
<file_name>MB_6.08_mod_CUDA_V10.exe</file_name>
</file_ref>
</app_version>

<app_version>
<app_name>setiathome_enhanced</app_name>
<version_num>608</version_num>
<flops>7400000000</flops>
<file_ref>
<file_name>AK_v8b_win_SSE3_AMD_GPU_CPU_team_V10.exe</file_name>
<main_program/>
</file_ref>
<file_ref>



Your advice is appreciated.

Dave

---

...
About CUDA performance - if the core where CUDA MB executed is occupied with some higher priority process (with priority higher than CUDA MB worker thread one) the app can experience substantional performance degradation. That's why CPU affinity mod was created (to give user with multicore CPUs ability to bound other processes on higher core numbers while leaving first core mostly for CUDA MB.) Other BOINC apps executed on even more low priority and should not interfere with CUDA MB.

Hmm..

It's only a crunching rig.. so nothing other on it..

So.. I guess.. it can be only the 25 % CPU / 100 % Core peaks from boinc.exe..
This could/would explain the CUDA performance lost?
This peaks are maybe 1 time/min. and last ~ 3 - 5 sec.

For example if I would crunch only on the GPUs this wouldn't help to have all the time 100 % GPU performance, or?

If I remember correct, in past as I crunched only on the GPUs - boinc.exe isn't also fixed only at a special CPU-Core.
Because in TaskManager all 4 CPU-Core showed usage-peaks as boinc.exe took 25 % CPU.

So finally I think my boinc.exe peaks couldn't reduce my CUDA performance, or?


What could be the cause?
It's all disabled (in BIOS and Windows) what I don't need.. only pure crunching rig.

---

@ FiveHamlet

Please have a look here in this thread:

app_info for AP500, AP503, MB603 and MB608

For the stock CUDA app you can take Raistmer's V10 or V11 CUDA app.
WITHOUT other things!
ONLY the 3 .dll's and the CUDA app of Raistmer.


EDIT:
If you take BOINC V6.6.20 you don't need the other parts of Raistmer's mods.

Message 884207

---

Thanks Sutaru
I have used the info file at the start of that thread and have it working ok on my i7 now.
I just have the AMD rig that I want to optomise.I thought that Raistmer's
apps may be better.I have it doing AP's at the moment cant get any Cuda.
I am running 6.6.20

---

Thanks Sutaru
I have used the info file at the start of that thread and have it working ok on my i7 now.
I just have the AMD rig that I want to optomise.I thought that Raistmer's
apps may be better.I have it doing AP's at the moment cant get any Cuda.
I am running 6.6.20

For your Intel i7 and AMD Phenom you can use Raistmer's V10 or V11 CUDA app with his .dll's.
Use also nearly the same app_info.xml .
You must take the opt. apps for your CPUs.
Max for MB: Intel - SSSE3x [SSE4 would work also - but AFAIK the SSE4 would run faster only on Core2 Duo] , AMD - SSE3 [both AK v8.0]
Max for AP: Intel and AMD - SSE3


You can find the newest available opt. CPU applications on the bottom of this thread:
http://setiathome.berkeley.edu/forum_thread.php?id=31810


Then change the app_info.xml [from the other thread] for your Intel and AMD with the names from the opt. CPU apps..


And then, it must work.. if not.. post.. :-)
[maybe there in the other thread]


I would recommend this opt. CUDA V10 app with opt. CPU CUDA start: directly download
In the forum from message #263


And the .dll's from the opening post of the thread.

You must register on the opt. crew webside for download.

---

I am trying to use V10 and 6.6.20 and couldn't get any CUDA either, only AP. I finally had to turn AP off and now I can't get any tasks at all.

I am trying to use V10 and 6.6.20 and couldn't get any CUDA either, only AP. I finally had to turn AP off and now I can't get any tasks at all.

I had similar problems. When I tried to use V10 team or V9 team with 6.6.20, my existing MB CUDA tasks would error out and I wouldn't get any more. Once I downgraded to V9 non-team everything worked perfectly.

---

You will be assimilated...bunghole!

@ elgar, Borgholio

Why it work for all the others and for me also? ;-)

If you like, follow my upper 'instructions' and with help of the app_info.xml-thread and with only Raistmer's CUDA V10 app and .dll's and give it a try.. :-)


..maybe it didn't worked because you took the complete mod?
I don't know..

If you made a mistake in your app_info.xml you wouldn't get CUDA tasks..

---