1) This "team" will do AP tasks, MB tasks and will utilize both CPU and GPU for MB tasks. AP tasks still only for CPU.

Hi,

I have now run V8 until it emptied my MB cache (not receiving MBs), but when the MB WU-cache is empty the GPU is switched for a CPU and 5 CPUs are used for AP - not 4. I know I can manually change ncpus, but that will have to be changed back when I receive MB WUs - not really optimal - or intentional...?

Morten

Sure it not optimal. But could you suggest something better - that's the question. Optimal is not to allow MB queue drain, but it's not always possible of course. But running 5 CPU apps on 4 cores more effective than to run 3 CPU app on 3 cores most time.

OK, so it's working as designed - no problemo! I'm gonna temporarily install BOINC 6.2.2 to get more MBs - (no WUs available now, though...)

Morten

---

Morten Ross

CUDA CONTROL ISSUES OR "HELP! I FELL UP AND I CANT GET DOWN."

CUDA and my GPU have taken over my P4 3.2gh + 2GB memory
I have a GEFORCE 8400 GS with 512MB. Occurs with both Boinc 5.4.5 and 6.6.2, I have also updated to the Nvidia xxx1882 drivers

Problems include cursor & page freeze - (program is working but over uses resources seriously)

Also, as mentioned here, programs programs do not exit.

Overall Software usage is set at the recommended 100% per program but processor and memory usage is set to less than 50%

I have also used the Preferences to dial back Boinc Usage to less than 50 percent of total resources and not to process while the computer is in use but the darn thing will not die or slow even when I "suspend"

I can refuse to process but nothing in between.

I know the answer is out there.

:>)

---

The welfare of the people is the highest law - Cicero

If no one complains, the people must be satisfied.

All smooth here.came just at the right time,I only have mbs' in my stash so have been running on cuda alone with all cores idle, now all 5 doing all those mbs I have. You shure are good at this stuff! Thanks again.

@ Stacey B.:

Boinc 6.6.2 gave me those problems too (uncontrolled workunit downloads, empty screens and vanishing Boinc Manager- but Boinc.exe still running and furiously downloading WUs, had to kill it from Task Manager> processes).

Boinc 6.6.0 does not have those problems, likewise Boinc 6.4.5 is fine.

I'm happily running Raistmer's version 8a optimized MultiBeam Cuda on my GeForce 8800GT/512mb card with nVidia 181.20 video drivers plus two CPU cores. As he said, I set ncpus = 3 = 2 CPU cores + 1 GPU in a cc_config.xml file.

---

Sabertooth Z77, i7-3770K@4.2GHz, GTX680, W8.1Pro x64
P5N32-E SLI, C2D E8400@3Ghz, GTX580, Win7SP1Pro x64 & PCLinuxOS2015 x64

CUDA CONTROL ISSUES OR "HELP! I FELL UP AND I CANT GET DOWN."

CUDA and my GPU have taken over my P4 3.2gh + 2GB memory
I have a GEFORCE 8400 GS with 512MB. Occurs with both Boinc 5.4.5 and 6.6.2, I have also updated to the Nvidia xxx1882 drivers

Problems include cursor & page freeze - (program is working but over uses resources seriously)

Also, as mentioned here, programs programs do not exit.

Overall Software usage is set at the recommended 100% per program but processor and memory usage is set to less than 50%

I have also used the Preferences to dial back Boinc Usage to less than 50 percent of total resources and not to process while the computer is in use but the darn thing will not die or slow even when I "suspend"

I can refuse to process but nothing in between.

I know the answer is out there.

:>)

SSE level of your P4 ? Did you read first message of this thread?

There is SSE3-level package available at Lunatics site.
So, from now recent (and not so recent) AMD chips are supported too.

There is SSE3-level package available at Lunatics site.
So, from now recent (and not so recent) AMD chips are supported too.

So far so good, thanks Raistmer and all involved in getting this out so fast.

---

Ohhhh you said applicable only to **SSSE3** in your first message of the thread, duh.

I thought SS3 . . . .

Ok, I will delete and install the correct software. Except for the freezing problem it was going quite well.


Thank you

---

The welfare of the people is the highest law - Cicero

If no one complains, the people must be satisfied.

Ok, my little C2D is chugging along nicely except for one little annoyance. During our last little outage I ran out of MBs. When it didn't have anything to keep the cuda busy it picked up a third AP to crunch. I let it go for awhile and noticed it caused both APs to slow down. Now that I have more MBs, it seems to think it was ok to run three APs so it keeps picking up on the AP instead of starting an MB on the Cuda. I have to watch the little bugger to make sure it picks up an MB each time it finishes one.

---

PROUD MEMBER OF Team Starfire World BOINC

Ok, my little C2D is chugging along nicely except for one little annoyance. During our last little outage I ran out of MBs. When it didn't have anything to keep the cuda busy it picked up a third AP to crunch. I let it go for awhile and noticed it caused both APs to slow down. Now that I have more MBs, it seems to think it was ok to run three APs so it keeps picking up on the AP instead of starting an MB on the Cuda. I have to watch the little bugger to make sure it picks up an MB each time it finishes one.

LOL, BOINC is quite capable of doing that itself without Raistmer's help.

Since adding CUDA to a quad (BOINC v6.4.5, ncpus=5), I'm often seeing five CPU cores in use, plus a sixth task on the GPU. Doesn't matter if you're running fast tasks on the GPU (sorts itself out with the reschedule when the GPU task finishes), but when you pick up the occasional VLAR it can stay in 'n+1+1' mode for quite a while.

Raistmer posted 26Jan09

V8a update available.
Changes:

- VLAR autokill mod enabled for CUDA app
- PGO redone for CPU app
- Wall-clock elapsed time since last restart (or since start of task) is added to stderr for both apps.

@Raistmer:
Could you provide an http link to files.mail.ru to download your optimized version 8a "Team work" package, for the benefit of those who don't have an account at Lunatics? Thanks.

---

Sabertooth Z77, i7-3770K@4.2GHz, GTX680, W8.1Pro x64
P5N32-E SLI, C2D E8400@3Ghz, GTX580, Win7SP1Pro x64 & PCLinuxOS2015 x64

@Raistmer:
Could you provide an http link to files.mail.ru to download your optimized version 8a "Team work" package, for the benefit of those who don't have an account at Lunatics? Thanks.

Sure: http://files.mail.ru/NXVFSE

BTW, SSE2 version coming soon

Ok, have a non issue but thought I'd report what I'm seeing.

In Windows Task Manager I see two instances of AK_v8 and one of the CUDA.exe

In Boinc Manager I see the correct amount of tasks runing for the settings I have in the cc config file for ncpus=2, which is one AK_v8 task and one CUDA.

---

Ok, have a non issue but thought I'd report what I'm seeing.

In Windows Task Manager I see two instances of AK_v8 and one of the CUDA.exe

In Boinc Manager I see the correct amount of tasks runing for the settings I have in the cc config file for ncpus=2, which is one AK_v8 task and one CUDA.

Yes, it just as it should be.
As I wrote in first message :
"
How it works:
for BOINC it looks as usual CPU opt app installed. BOINC will call CPU app (AK_v8b_win_SSSE3x_GPU_CPU_team.exe in our case) and assign one of SETI MB tasks for it. But this app aware of possibility to use GPU for computations. It will check if another instance (it knows only itself and its clones, so - no other GPU-related projects please) already use GPU and if not - will start GPU-related app (MB_6.08_mod_CPU_team_CUDA.exe in our case) and suspend itself until GPU app finish. This CUDA app will do all work as usually but will do it on GPU leave CPU almost free.
"
So on single core you should see one working AK_v8, one suspended AK_v8 (0 CPU time and 0 % CPU usage) and one CUDA app.

Oh ok, I think I get it now. sorry I didn't understand the first message So the second AK instance is what is feeding the gpu?

Edit: Ok I see now from your edit.

---

I got a question I cannot get Raistmer's app to work it freezes up BOINC and it ruins the WU's. I run 6.2.2. What do I have to do to get it to work?

@Paul:

I and many others had significant problems with Boinc 6.6.2 - see multiple other threads. My guess 6.6.2 is the root of your problems.

I reverted from 6.6.2 to 6.6.0 (and Raistmer said 6.4.5 is OK too), and I'm running Raistmer's version 8a "Team Work" optimized app that will crunch MB work on CPU&GPU concurrently and AstroPulse work. See this message to download optimized app package, and read his directions in that thread. It's quite easy to install & setup, and works great (be sure you're using 181.20 or later nVidia driver if your want to run MB Cuda on a Cuda-capable nVidia 8xxx or later video card).

---

Sabertooth Z77, i7-3770K@4.2GHz, GTX680, W8.1Pro x64
P5N32-E SLI, C2D E8400@3Ghz, GTX580, Win7SP1Pro x64 & PCLinuxOS2015 x64

Thank you so much for this!! Installed it on three machines. So far they are working beautifully. Interesting to see how my rac will compares to your previous AP/MB package I was useing. Keep up the great work!

Here are the hosts:
For various reasons hadn't got in a full 24 hour run yet but hopefull this new app willmake that possible when the project has more work.

5000+
6000+
PentiumD 2.80

---

"The most exciting phrase to hear in science, the one that heralds new discoveries, is not Eureka! (I found it!) but rather, 'hmm... that's funny...'" -- Isaac Asimov

@Paul:

I and many others had significant problems with Boinc 6.6.2 - see multiple other threads. My guess 6.6.2 is the root of your problems.

I reverted from 6.6.2 to 6.6.0 (and Raistmer said 6.4.5 is OK too), and I'm running Raistmer's version 8a "Team Work" optimized app that will crunch MB work on CPU&GPU concurrently and AstroPulse work. See this message to download optimized app package, and read his directions in that thread. It's quite easy to install & setup, and works great (be sure you're using 181.20 or later nVidia driver if your want to run MB Cuda on a Cuda-capable nVidia 8xxx or later video card).

Thanks I downloaded 6.4.5 and installed.
I downloaded Raistmer's 8a app but in the instruction it says

"4) <ncpus>NUMBER_OF_CORES+1</ncpus> is REQUIRED for productive work. If you let BOINC manage CPU cores number you will end up with one idle core, trust me ;)"

How do I do that?

@Paul:

Further above in this thread, Raistmer said:

6) For best CPU and GPU usage I recommend to set number of processors available for BOINC to real_number_of_cores+1. This will mitigate current BOINC bug with CPU+CUDA scheduling and will allow fully load CPU and GPU.
Here is example of minimal cc_config.xml file you need

<cc_config>
<options>
<ncpus>enter_number_of_cores+1_value_here</ncpus>
</options>
</cc_config>

You should put it in BOINCdata folder.

Create a new file using Notepad using the lines above (copy and paste), replace "enter_number_of_cores+1" with the integer value applicable to your CPU (thus use 2 if single core, 3 if dual core, 5 if quad core), then in Notepad choose Save As with filename "cc_config.xml" ensuring Save as type is "All files" (so that you save it with .xml extension, not as .txt). Put that cc_config.xml file in your Boinc folder (default folder path is C:\Documents and Settings\All Users\Application Data\BOINC).

Post back if you have any other questions, or to tell us how it works for you.

---

Sabertooth Z77, i7-3770K@4.2GHz, GTX680, W8.1Pro x64
P5N32-E SLI, C2D E8400@3Ghz, GTX580, Win7SP1Pro x64 & PCLinuxOS2015 x64