milkyway_separation 1.46 Linux x86_64 double OpenCL

Can somebody tell me what the 'short name' is for this app or where to find it? I'm after the name that goes into an app_config.xml file.

Thanx, Skip

As long as I'm begging...

Anybody want to share an app_config.xml for MW roughly appropriate for a 8 core CPU box with an AMD RX 580 card? I've next to no experience with these.

Again Thanx, Skip

You can always look in your client_state.xml file for the app_name.

---

You can try,

<app_config>
<app>
<name>milkyway</name>
<gpu_versions>
<gpu_usage>0.25</gpu_usage>
<cpu_usage>0.25</cpu_usage>
</gpu_versions>
</app>
</app_config>

As long as I'm begging...

Anybody want to share an app_config.xml for MW roughly appropriate for a 8 core CPU box with an AMD RX 580 card? I've next to no experience with these.

Again Thanx, Skip

This is the one I use:

<app_config>

<app>
<name>milkyway</name>
<gpu_versions>
<gpu_usage>0.5</gpu_usage>
<cpu_usage>0.5</cpu_usage>
</gpu_versions>
</app>
<app_version>
<app_name>milkyway_nbody</app_name>
<max_concurrent>5</max_concurrent>
<plan_class>mt</plan_class>
<avg_ncpus>2</avg_ncpus>
<cmdline>--nthreads 2</cmdline>
</app_version>
</app_config>

I do delete the top section if I'm not running gpu tasks. The top gpu part tells your pc to run 2 gpu tasks at the same time while the bottom cpu part says to run 5 tasks at one time each using 2 cpu cores.

client_state.xml

Only if you've executed it before. Which in this case I had so should've looked there.

As long as I'm begging...

Anybody want to share an app_config.xml for MW roughly appropriate for a 8 core CPU box with an AMD RX 580 card? I've next to no experience with these.

Again Thanx, Skip

This is the one I use:

<app_config>

<app>
<name>milkyway</name>
<gpu_versions>
<gpu_usage>0.5</gpu_usage>
<cpu_usage>0.5</cpu_usage>
</gpu_versions>
</app>
<app_version>
<app_name>milkyway_nbody</app_name>
<max_concurrent>5</max_concurrent>
<plan_class>mt</plan_class>
<avg_ncpus>2</avg_ncpus>
<cmdline>--nthreads 2</cmdline>
</app_version>
</app_config>

I do delete the top section if I'm not running gpu tasks. The top gpu part tells your pc to run 2 gpu tasks at the same time while the bottom cpu part says to run 5 tasks at one time each using 2 cpu cores.

Thanx these are a big help!

OK, running a bunch on the GPU right now and they seem to working two at a time even though I pretty sure these were in the que B4 I made the changes to the app_config. I did a "re-read config files" and they seemed to start running 2 at a time even though the line in boinc manager still says
"Ready to start (0.954 CPUs + 1 AMD/ATI GPU". Didn't realize it would change how they would execute on the fly w/o downloading new ones.

Anyway thanx to all. Will continue to tweek but right now I gotta go see if I can the proper driver/OpenCL from AMD installed for an OLD Oland card on a computer out in the garage.

Here's what I'm running with on this [url]https://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=979753 [/url] box right now:

<app_config>
 <project_max_concurrent>10</project_max_concurrent>

 <app>
  <name>milkyway_nbody</name>
  <max_concurrent>4</max_concurrent>
   <app_version>
    <app_name>milkyway_nbody</app_name>
    <plan_class>mt</plan_class>
    <avg_ncpus>2</avg_ncpus>
    <cmdline>--nthreads 2</cmdline>
   </app_version>
 </app>

 <app>
  <name>milkyway</name>
  <gpu_versions>
   <gpu_usage>0.5</gpu_usage>
   <cpu_usage>0.5</cpu_usage>
  </gpu_versions>
 </app>
</app_config>

Thanx all for the SUPER support.

OK, running a bunch on the GPU right now and they seem to working two at a time even though I pretty sure these were in the que B4 I made the changes to the app_config. I did a "re-read config files" and they seemed to start running 2 at a time even though the line in boinc manager still says
"Ready to start (0.954 CPUs + 1 AMD/ATI GPU". Didn't realize it would change how they would execute on the fly w/o downloading new ones.

Anyway thanx to all. Will continue to tweek but right now I gotta go see if I can the proper driver/OpenCL from AMD installed for an OLD Oland card on a computer out in the garage.

Here's what I'm running with on this [url]https://milkyway.cs.rpi.edu/milkyway/show_host_detail.php?hostid=979753 [/url] box right now:

<app_config>
 <project_max_concurrent>10</project_max_concurrent>

 <app>
  <name>milkyway_nbody</name>
  <max_concurrent>4</max_concurrent>
   <app_version>
    <app_name>milkyway_nbody</app_name>
    <plan_class>mt</plan_class>
    <avg_ncpus>2</avg_ncpus>
    <cmdline>--nthreads 2</cmdline>
   </app_version>
 </app>

 <app>
  <name>milkyway</name>
  <gpu_versions>
   <gpu_usage>0.5</gpu_usage>
   <cpu_usage>0.5</cpu_usage>
  </gpu_versions>
 </app>
</app_config>

Thanx all for the SUPER support.

Anytime!!

Only new tasks downloaded after applying the app_config.xml change will be identified with the changes implemented by your app_config. The pre-existing tasks in your cache will still be indentifed by the conditions that existed when they were downloaded.

As you found out, reading the config files applies the changes immediately in how BOINC runs.

---