Howdy!

Have a couple questions.... especially where a Ryzen 9 5950X is concerned...


1) Why are N-Body tasks limited to 16 CPUs?

2) Why can't we set how many CPUs N-Body uses?

8-)

Only the project can tell you why the limit. But you can run on any number of virtual cores (threads) up to that limit using an "app_config.xml" file.
This is for running on 12 cores, with one work unit per core; change as desired.

<app_config>
<app>
<name>milkyway_nbody</name>
<max_concurrent>12</max_concurrent>
</app>

<app_version>
<app_name>milkyway_nbody</app_name>
<plan_class>mt</plan_class>
<avg_ncpus>1</avg_ncpus>
<cmdline>--nthreads 1</cmdline>
</app_version>
</app_config>

You create it in a text processor (Notepad), use the "save as" function to save it as an ".xml" file, and place it in the project data folder.
Then restart BOINC or read in the file to activate it.
I usually just reboot. It works on Windows or Linux.

Worked well for me with minor mod!

THANKS!

8-)

Long Runner aborted after 4 days 10 hours when I finally spotted it.

de_nbody_06_16_2021_v176_40k__data__7_1626375792_418770

Thanks
Bill F

---

In October of 1969 I took an oath to support and defend the Constitution of the United States against all enemies, foreign and domestic;
There was no expiration date.

Sounds like you had the problem reported here. Next time before aborting try restarting BOINC.

I installed my app_config and made sure I read configs before switching to All New Work. But it keeps sending me 16c WUs, hundreds of them. I don't know what to do except Abort them. Be nice if we had more control over our work. E.g, there's no way to specify number of CPUs per WU in Edit My Preferences. Nor can one get n-body CPU WUs and just GPU WUs since we can get one of the other or all three.

<app_config>
<app>
    <name>milkyway</name>
    <gpu_versions>
        <cpu_usage>1.0</cpu_usage>
        <gpu_usage>1.0</gpu_usage>
    </gpu_versions>
    <max_concurrent>4</max_concurrent>
</app>
<app>
    <name>milkyway_nbody</name>
    <plan_class>mt</plan_class>
    <avg_ncpus>1</avg_ncpus>
    <cmdline>--nthreads 1</cmdline>
    <max_concurrent>1</max_concurrent>
</app>
</app_config>

From my client_state:

<app_version>
    <app_name>milkyway_nbody</app_name>
    <version_num>182</version_num>
    <platform>x86_64-pc-linux-gnu</platform>
    <avg_ncpus>16.000000</avg_ncpus>
    <flops>89728291437.088242</flops>
    <plan_class>mt</plan_class>
    <api_version>7.6.33</api_version>
    <file_ref>
        <file_name>milkyway_nbody_1.82_x86_64-pc-linux-gnu__mt</file_name>
        <main_program/>
    </file_ref>
</app_version>

It should be:

<avg_ncpus>1.000000</avg_ncpus>

I checked and they actually running with 16 CPU threads so it's not just reporting ti wrong.

Now they're coming as 16C but not displaying that they're 16C.

First question to ask is did you preserve the file type as an .xml file? Make sure you didn't create it as app_config.xml.txt.

BOINC will ignore the file if not named correctly.

Second question is do you see the app_config being read for Einstein in the Event Log at startup?

If you don't see the file being recognized by BOINC it isn't being acted on and is being ignored.

Both cases would explain why your tasks are still being sent as 16c tasks. Also, any old tasks in your cache before you made the change will still show as 16c. All new tasks will have the correct intended parameters.

---

First question to ask is did you preserve the file type as an .xml file? Make sure you didn't create it as app_config.xml.txt.
BOINC will ignore the file if not named correctly.

That sounds like a Windoze problem. I create and/or edit them remotely using:
sudo xed /var/lib/boinc-client/projects/milkyway.cs.rpi.edu_milkyway/app_config.xml

Second question is do you see the app_config being read for Einstein in the Event Log at startup?
If you don't see the file being recognized by BOINC it isn't being acted on and is being ignored.

Do you mean this line from the BoincTasks Message tab?
Rig-01 137251 Milkyway@Home 2/19/2022 10:06:53 PM Found app_config.xml

Both cases would explain why your tasks are still being sent as 16c tasks. Also, any old tasks in your cache before you made the change will still show as 16c. All new tasks will have the correct intended parameters.

Yea I thought maybe I wasn't clearing the decks before running the edited app_config.xml but I just ran a test again (from having no MW WUs & reading config files) and the same thing happened.
I have 2 computers where it works correctly and I've been comparing to ones that keep getting 16Cs and I can find no difference. I looked for typos in my syntax, permissions, and ownership. I'm wondering if it requires a BOINC restart (sudo /etc/init.d/boinc-client restart) to get <avg_ncpus>1.000000</avg_ncpus> written into the client_state.xml file. Problem is I'm running an eclectic mix: GAIA with no checkpointing, ARP1 (_2s, _3s & _4s) with checkpointing every few hours and QuChemPedia with checkpointing every 30 seconds or who knows when.

Are these bona vide multithreaded WUs or just 16 WUs in one wrapper that are extremely inefficient???
Up to 4 threads and I wouldn't even care but 16, yikes!!! Control of the number of threads per CPU WU could be placed in our MW Preferences if they wanted to.

Both of the computers that have been running 1C nbody WUs all day have now reverted to DLing 16C WUs.
It cannot be something I'm doing.
These WUs are misconfigured.