Here is the link for working with app_info.xml files.
http://www.xtremesystems.org/forums/showthrea...09-Working-app_info-files

You can add other projects but the app_info file must be coded for this with an entry for each additional project you wish to run.

The file goes in c:\program data\boinc\projects\www.worldcommunitygrid.org directory - assuming that you are using the default boinc directories. If not, adjust accordingly.

As long as you have been crunching at least one ver 7.05 HCC wu, you should have the science app needed for crunching the HCC-GPU wu's. If so, you can exit boinc manager making sure that you have also shut down all of the running science apps, put the xml file in the above noted directory and then restart boinc.

You should set the max and avg cpu parameters to 1.0 and the count parameter to 1 divided by the number of gpu threads you want. For example, here is my xml for a 2600k running 2 7950's with 8 threads per gpu.

<app_info>

<app>
<name>hcc1</name>
<user_friendly_name>Help Conquer Cancer</user_friendly_name>
</app>
<file_info>
<name>wcg_hcc1_img_7.05_windows_intelx86__ati_hcc1</name>
<executable/>
</file_info>
<file_info>
<name>hcckernel.cl.7.05</name>
<executable/>
</file_info>
<app_version>
<app_name>hcc1</app_name>
<version_num>705</version_num>
<platform>windows_intelx86</platform>
<plan_class>ati_hcc1</plan_class>
<avg_ncpus>0.5</avg_ncpus>
<max_ncpus>1.0</max_ncpus>
<coproc>
<type>ATI</type>
<count>0.125</count>
</coproc>
<file_ref>
<file_name>wcg_hcc1_img_7.05_windows_intelx86__ati_hcc1</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>hcckernel.cl.7.05</file_name>
<open_name>hcckernel.cl</open_name>
</file_ref>
</app_version>
</app_info>

As you can see, I have the avg cpu parm set to .5 since I have 2 gpu's. You would use 1.0 if you only have one gpu.

I have the count set to 0.125 since 1/8(threads) is 0.125

edit:note: program data is a hidden directory - you will need to unhide that. Let us know if you need help there.

edit2: the number of threads depends of various factors but as a general rule should not exceed the number of thread your cpu can run. i have a 2600k which runs 4 cores and 8 threads so I have selected a count of 8 - but that's 8 for each gpu or a total of 16.

Hi themoonscrescent,

1. You have to install at least 7.0.42 Boinc Manager / Boinc Client from here (try 7.0.44): http://boinc.berkeley.edu/dev/forum_thread.php?id=6698

2. Here is my app_config.xml (containing all projects except of CEP2) which has to be modified by you according to your preferences and placed in C - ProgramData (hidden by default) - BOINC - projects - www.worldcommunitygrid.org (if you are running W7 64b, for other OS's follow SekeRob's recommendation)

<app_config>
<app>
<name>hcc1</name>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>dsfl</name>
<max_concurrent>8</max_concurrent>
</app>
<app>
<name>dddt2</name>
<max_concurrent>8</max_concurrent>
</app>
<app>
<name>sn2s</name>
<max_concurrent>8</max_concurrent>
</app>
<app>
<name>gfam</name>
<max_concurrent>8</max_concurrent>
</app>
<app>
<name>c4cw</name>
<max_concurrent>8</max_concurrent>
</app>
<app>
<name>faah</name>
<max_concurrent>8</max_concurrent>
</app>
<app>
<name>hpf2</name>
<max_concurrent>8</max_concurrent>
</app>
<app>
<name>beta</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta3</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta5</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta7</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta10</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta11</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta12</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta13</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta14</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta15</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta17</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
<app>
<name>beta18</name>
<max_concurrent>17</max_concurrent>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app>
</app_config>

The part <app>
<name>hcc1</name>
<gpu_versions>
<gpu_usage>.083333</gpu_usage>
<cpu_usage>.25</cpu_usage>
</gpu_versions>
</app> allows me to run 12 HCC1 GPU tasks (on one GPU: 1/12 = 0.08333) using 3 CPU cores (4 tasks on each CPU core: 1/4 = 0.25) and choose what CPU projects will be run on the remaining 5 CPU threads (I have octo CPU) with Profile settings I described here http://www.worldcommunitygrid.org/forums/wcg/...d,34395_offset,100#405515

If you wish to run e.g. 3 C4CW tasks + 2 FAAH tasks, keep in your app_config only

<app_config>
<app>
<name>c4cw</name>
<max_concurrent>3</max_concurrent>
</app>
<app>
<name>faah</name>
<max_concurrent>2</max_concurrent>
</app>
</app_config>

I added all those "BetaXX" just to check if they will work in case WCG will run it in the future based on info from uplinger http://www.worldcommunitygrid.org/forums/wcg/...ad,34480_offset,50#407581 (and read also next post by SekeRob in the same thread)

3. If it will not work for you, post the details about your system and projects you would like to run and we will do our best to help you

Good luck and cheers

Config menus? The Advanced menu in BOINC Manager has an option to read in whatever you entered in the cc_config.xml. The plan is to do same for the app_config.xml, which *replaces* app_info. It has 2 simple lines to set the fractions a GPU task uses of GPU and CPU core, to multithread your GPU. Whilst, reading in the config files does not effectuate all changes, often still requiring to restart the client... it's like good old windows getting patched... still requiring a boot most of the time. KSplice for Linux the MS Bros still have not figured out.

Exactly. All you need it the correct app name. You can run any project and dedicate any amount of CPU cores or GPU resources you want to each project by creating an app_config file for the projects you want to run and placing it in the correct project folder.
I was running 4 WCG, 2 SIMAP and 2 SkynetPOGS CPU tasks on an i7-860 by doing what I stated above. I then switched it to 8 WCG GPU tasks (0.50 CPU cores per task), 2 SIMAP and 2 POGS. Worked flawlessly. The combos are endless and the file creation is infinitely easier.

Not cc_config.xml, the new file is called app_config.xml


Yes, for most users just adding info to run multiple HCC1 on their GPU at once is enough, while all CPU-projects works as before.

Only if WCG releases a new XYZ-project for GPU would it be neccessary to add another application to app_config.xml for most users.