Just upgraded to the GTX 660ti, great card for the $$$. Was excited to fire it up and see how fast it crunched the opencl units. After getting everything up and running and observing for quite a while, I noticed the GPU utilization was maxxed at about 60%. Having had success getting multiple GPU wu's to run for SETI, I figured I'd give it a go with Milkyway. I achieved success (now have 2 WU's crunching simultaneously @ 100% GPU usage) by using the following app_info.xml file in the MilkyWay project folder. My problem currently is that without an application profile for the standard CPU app, the client doesn't call out for or run any of the standard CPU WU's. Was wondering if any of the really smart folks out here could help me with updating my app_info.xml. I am running 7.0.42 version of BOINC on Win7 X64

app_info.xml Contents follow:

<app_info>

<app>
<name>milkyway</name>
</app>

<file_info>
<name>milkyway_separation_1.02_windows_x86_64__opencl_nvidia.exe</name>
<executable/>
</file_info>

<app_version>
<app_name>milkyway</app_name>
<version_num>102</version_num>
<avg_ncpus>0.05</avg_ncpus>
<max_ncpus>1</max_ncpus>

<coproc>
<type>CUDA</type>
<count>.45</count>
</coproc>
<cmdline></cmdline>
<file_ref>
<file_name>milkyway_separation_1.02_windows_x86_64__opencl_nvidia.exe</file_name>
<main_program/>
</file_ref>
</app_version>

</app_info>

In BOINC 7.0.42 you can use app_config.xml in stead of app_info.xml.
Set the Milkyway-project to No New Work (NNW), finish all your current Milkyway-tasks.
Stop BOINC.
Delete your app_info.xml (first backup a copy!) and create the file app_config.xml in the same directory with content:

<app_config>
	<app>
		<name>milkyway</name>
		<gpu_versions>
			<gpu_usage>0.5</gpu_usage>
			<cpu_usage>0.05</cpu_usage>
		</gpu_versions>
	</app>
</app_config>

Start BOINC.
This should allow crunching CPU-tasks if you have selected those in your Milkyway-preferences.

Personally I don't crunch Milkyway on the CPU. The GPU does it much more efficient, so i have not tested the above.

So please report back if the above app_config.xml allows the CPU-crunching. If not: ask some other person for the extra CPU-lines needed in the app_info.xml.

Edit: the complete app_info.xml for CPU-crunching is already in another thread: http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3064&nowrap=true#56288

TeeVeeEss,

Thank you for your reply, it pointed me in the right direction. I now have 8 threads of CPU work on my I-7 3770K, and two WU's going on my GTX 660TI. For those interested, crunching a single WU on the 660 took about 17 minutes. Crunching 2 at a time takes a little longer (22mins) but produces a higher hourly output and keeps the 660 at near 100% utilization.

Here is the app_info.xml file that made it work:

<app_info>
<app>
<name>milkyway</name>
</app>
<file_info>
<name>milkyway_separation_1.00_windows_x86_64.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>milkyway</app_name>
<version_num>100</version_num>
<cmdline></cmdline>
<file_ref>
<file_name>milkyway_separation_1.00_windows_x86_64.exe</file_name>
<main_program/>
</file_ref>
</app_version>

<app>
<name>milkyway_nbody</name>
</app>
<file_info>
<name>milkyway_nbody_1.04_windows_x86_64__mt.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>milkyway_nbody</app_name>
<version_num>104</version_num>
<cmdline></cmdline>
<file_ref>
<file_name>milkyway_nbody_1.04_windows_x86_64__mt.exe</file_name>
<main_program/>
</file_ref>
</app_version>

<app>
<name>milkyway</name>
</app>
<file_info>
<name>milkyway_separation_1.02_windows_x86_64__opencl_nvidia.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>milkyway</app_name>
<version_num>102</version_num>
<flops>2.0e11</flops>
<avg_ncpus>0.05</avg_ncpus>
<max_ncpus>1</max_ncpus>
<coproc>
<type>CUDA</type>
<count>0.45</count>
</coproc>
<cmdline></cmdline>
<file_ref>
<file_name>milkyway_separation_1.02_windows_x86_64__opencl_nvidia.exe</file_name>
<main_program/>
</file_ref>
</app_version>
</app_info>

Update: Though the app_info.xml gave good indications initially, all of the milkyway_nbody104 tasks resulted in computation error. 102 units (opencl) are completing just fine, and i'm waiting for the first few 100 separation units to finish.

The 1.00 separation CPU tasks are completing normally. I'll be removing the portion of the file applicable to nbody104 units which will make the file look like this:

<app_info>
<app>
<name>milkyway</name>
</app>
<file_info>
<name>milkyway_separation_1.00_windows_x86_64.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>milkyway</app_name>
<version_num>100</version_num>
<cmdline></cmdline>
<file_ref>
<file_name>milkyway_separation_1.00_windows_x86_64.exe</file_name>
<main_program/>
</file_ref>
</app_version>

<app>
<name>milkyway</name>
</app>
<file_info>
<name>milkyway_separation_1.02_windows_x86_64__opencl_nvidia.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>milkyway</app_name>
<version_num>102</version_num>
<flops>2.0e11</flops>
<avg_ncpus>0.05</avg_ncpus>
<max_ncpus>1</max_ncpus>
<coproc>
<type>CUDA</type>
<count>0.45</count>
</coproc>
<cmdline></cmdline>
<file_ref>
<file_name>milkyway_separation_1.02_windows_x86_64__opencl_nvidia.exe</file_name>
<main_program/>
</file_ref>
</app_version>
</app_info>

Here is the app_info.xml file that made it work:

<app_info>
<app>
<name>milkyway</name>
</app>
<file_info>
<name>milkyway_separation_1.00_windows_x86_64.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>milkyway</app_name>
<version_num>100</version_num>
<cmdline></cmdline>
<file_ref>
<file_name>milkyway_separation_1.00_windows_x86_64.exe</file_name>
<main_program/>
</file_ref>
</app_version>

<app>
<name>milkyway_nbody</name>
</app>
<file_info>
<name>milkyway_nbody_1.04_windows_x86_64__mt.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>milkyway_nbody</app_name>
<version_num>104</version_num>
<cmdline></cmdline>
<file_ref>
<file_name>milkyway_nbody_1.04_windows_x86_64__mt.exe</file_name>
<main_program/>
</file_ref>
</app_version>

<app>
<name>milkyway</name>
</app>
<file_info>
<name>milkyway_separation_1.02_windows_x86_64__opencl_nvidia.exe</name>
<executable/>
</file_info>
<app_version>
<app_name>milkyway</app_name>
<version_num>102</version_num>
<flops>2.0e11</flops>
<avg_ncpus>0.05</avg_ncpus>
<max_ncpus>1</max_ncpus>
<coproc>
<type>CUDA</type>
<count>0.45</count>
</coproc>
<cmdline></cmdline>
<file_ref>
<file_name>milkyway_separation_1.02_windows_x86_64__opencl_nvidia.exe</file_name>
<main_program/>
</file_ref>
</app_version>
</app_info>

Your nbody section is missing the reference to libgomp_64-1_nbody_1.04.dll and pthreadGC2_64_nbody_1.04.dll. And while you are at it, you should add the mt plan class. Additionally I would define the number of threads to use too; it was helping in the past to force BOINC to make use of multithreading.

<app>
<name>milkyway_nbody</name>
</app>
<file_info>
<name>milkyway_nbody_1.04_windows_x86_64__mt.exe</name>
<executable/>
</file_info>
<file_info>
<name>libgomp_64-1_nbody_1.04.dll</name>
<executable/>
</file_info>
<file_info>
<name>pthreadGC2_64_nbody_1.04.dll</name>
<executable/>
</file_info>
<app_version>
<app_name>milkyway_nbody</app_name>
<version_num>104</version_num>
<plan_class>mt</plan_class>
<!-- the following 3 lines are setting 4 threads per WU -->
<avg_ncpus>4</avg_ncpus>
<max_ncpus>4</max_ncpus>
<cmdline>--nthreads=4 --disable-opencl</cmdline>
<file_ref>
<file_name>milkyway_nbody_1.04_windows_x86_64__mt.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>libgomp_64-1_nbody_1.04.dll</file_name>
<open_name>libgomp_64-1.dll</open_name>
<copy_file/>
</file_ref>
<file_ref>
<file_name>pthreadGC2_64_nbody_1.04.dll</file_name>
<open_name>pthreadGC2_64.dll</open_name>
<copy_file/>
</file_ref>
</app_version>

You might try it with and without the --disable-opencl switch.

hey folks - I bought a nvidia gtx 650 and have a asusp5g41T Mb witha Q6600 ddr3 ram [8 gig ]and win 7 . I want to get the video card to run over 57 % [90 + maybe] . but I have no clue how to mod the program . I use the MB video the ram runs 25 -33 % useage and is very fast the cpu runs 2.7 ish , running win 7 OS gpuz sees 1.1 pcie too . so I think the video card is throttled back the slot

hey folks - I bought a nvidia gtx 650 and have a asusp5g41T Mb witha Q6600 ddr3 ram [8 gig ]and win 7 . I want to get the video card to run over 57 % [90 + maybe] . but I have no clue how to mod the program . I use the MB video the ram runs 25 -33 % useage and is very fast the cpu runs 2.7 ish , running win 7 OS gpuz sees 1.1 pcie too . so I think the video card is throttled back the slot

Easy Peasy...1st upgrade to a Boinc version of 7.0.40 or newer from here:
[url]http://boinc.berkeley.edu/dl/?C=M;O=D[/url[

2nd use the app_config.xml file that was posted a few messages ago. Then you will be running two units at once on your gpu!

TeeVeeEss,

Thank you for your reply, it pointed me in the right direction. I now have 8 threads of CPU work on my I-7 3770K, and two WU's going on my GTX 660TI. For those interested, crunching a single WU on the 660 took about 17 minutes. Crunching 2 at a time takes a little longer (22mins) but produces a higher hourly output and keeps the 660 at near 100% utilization.

The problems with an app_info file are as you are finding out, the file version numbers. IF you were to switch to an app_config file then you would never have to worry about the project changing file versions again. The app_info file requires you to have all the file versions and file names correct, the app_config file doesn't care about all that stuff, it just works.

well I open the Boinc folder and have no idea where to look [ first time ] I have the newer program too . any pointers . never operated on the programs just let them run

BTW newegg has the 650 ti on sale

well I open the Boinc folder and have no idea where to look [ first time ] I have the newer program too . any pointers . never operated on the programs just let them run

The Boinc folder I mean is in a hidden folder called Program Data, to get to it open Window Explorer then press the alt key and then click tools, folder options, tools and then I think it is the 7th check box down check it to show hidden files and folders. Say yes to any are you sure notices and then you should be able to see the folder.

Make SURE you ONLY use Notepad to make and save the app_config.xml file. Then when the file is in place stop and exit Boinc and restart it. You shouldn't even lose any files like the app_info.xml file does.

thanks I going to try next week , also I have my power settings cranked up . are there any theyt will keep my slot throbbed back ?

thanks I going to try next week , also I have my power settings cranked up . are there any theyt will keep my slot throbbed back ?

Now that is one thing I don't do, overclocking or underclocking has never 'stuck' for me. I used to oc the settings but they would always go back to the defaults, even if they were saved, so often that I stopped.