I have come to the conclusion that Moo will not run GPU tasks under any BOINC version above 6.12.34. I have tried both version 7.0.3 and 7.0.20 with no luck. I have over 290 Tasks just sitting and not starting. I have to run 7.X.X because 6.12.34 dose not allow me to set aside 1 GPU core for general computing, and the screen lag is unbearable.

---

What you do today you will have to live with tonight

I have come to the conclusion that Moo will not run GPU tasks under any BOINC version above 6.12.34.

You can get Moo to run under 7.x.x by using the app_info.xml file in the Moo Project folder. Use the count 1 option as listed below. Be sure to watch the temp on the GPU as this has usually increased the temps on my GPUs upward of 10C. I take it you're using the cc_config.xml file with the exclude GPU statement.

<app_info>
<app>
<name>dnetc</name>
<user_friendly_name>Distributed.net Client</user_friendly_name>
</app>
<file_info>
<name>dnetc_wrapper_1.3_windows_intelx86__ati14.exe</name>
<executable/>
</file_info>
<file_info>
<name>dnetc518-win32-x86-stream.exe</name>
<executable/>
</file_info>
<file_info>
<name>dnetc-gpu-1.3.ini</name>
</file_info>
<file_info>
<name>job-ati14-1.00.xml</name>
</file_info>
<app_version>
<app_name>dnetc</app_name>
<version_num>103</version_num>
<platform>windows_intelx86</platform>
<avg_ncpus>0.15</avg_ncpus>
<max_ncpus>1.0</max_ncpus>
<plan_class>ati14</plan_class>
<flops>1157115231469.729200</flops>
<file_ref>
<file_name>dnetc_wrapper_1.3_windows_intelx86__ati14.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>dnetc518-win32-x86-stream.exe</file_name>
<copy_file/>
</file_ref>
<file_ref>
<file_name>dnetc-gpu-1.3.ini</file_name>
<open_name>dnetc.ini</open_name>
<copy_file/>
</file_ref>
<file_ref>
<file_name>job-ati14-1.00.xml</file_name>
<open_name>job.xml</open_name>
<copy_file/>
</file_ref>
<coproc>
<type>ATI</type>
<count>1.0</count>
</coproc>
</app_version>
</app_info>

Yes, I have tried every way I can think of to get 6.12.34 to leave one GPU core out but nothing has worked.
Thanks I will try that app_info and see how it works.

---

What you do today you will have to live with tonight

Why don't run Moo with Boinc 7.x.x.
I don't wanna change back to an older version.
Is there a way to go with the 7.x.x.
Greetings

I am using 7.0.25 on W7-64 and Moo! has not skipped a beat.

I am using 7.0.25 on W7-64 and Moo! has not skipped a beat.

Same here, running great!
I do not see why K2 and SE are not able to run. I would ask if SE can run on 1 gpu (taking one out) mikey has/is using an cc_conifg.xml on his d-gpu set ups. SE Q-gpus, try d-gpus or start with 1 gpu then add as you go.

---

I am using 7.0.25 on W7-64 and Moo! has not skipped a beat.

Same here, running great!
I do not see why K2 and SE are not able to run. I would ask if SE can run on 1 gpu (taking one out) mikey has/is using an cc_conifg.xml on his d-gpu set ups. SE Q-gpus, try d-gpus or start with 1 gpu then add as you go.

Mine STOPPED running only one Moo unit though, it will work fine for other projects but Moo just doesn't do it anymore! I am using the exclude lines in cc_config.xml though NOT the app_info.xml file. Since it works for other projects I am wondering if it is the file name that is wrong, I have not played with it to figure that out yet. Here is what I have now that does NOT work:

<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
<exclude_gpu>
<url>http://http:/moowrap.net/</url>
<device_num>0</device_num>
</exclude_gpu>
<exclude_gpu>
<url>http://boinc.thesonntags.com/collatz/</url>
<device_num>1</device_num>
</exclude_gpu>
<report_results_immediately>1</report_results_immediately>
</options>
</cc_config>

It will run ONE Collatz unit or one whatever else unit, but NOT one Moo unit!

Been saying that for quite some time, Boinc 7.x.x and Moo dont get along other then for those who play around with files. Not right, should work out of the box. Not every one has the time or the inclination or the knowledge to play around with files.
Good luck with the roll back because once on to Boinc 7.x.x it starts behaving like a virus(other then folks with the knowledge) you may have to do a hard format. :)Like i said, not every one has the time or the inclination to dive under the HOOD. Wonder what Ageless has to say about this. Got quite excited when i brought it up on the Boinc Forums some where in Jan.
Regards.

Been saying that for quite some time, Boinc 7.x.x and Moo dont get along other then for those who play around with files. Not right, should work out of the box. Not every one has the time or the inclination or the knowledge to play around with files.
Good luck with the roll back because once on to Boinc 7.x.x it starts behaving like a virus(other then folks with the knowledge) you may have to do a hard format. :)Like i said, not every one has the time or the inclination to dive under the HOOD. Wonder what Ageless has to say about this. Got quite excited when i brought it up on the Boinc Forums some where in Jan.
Regards.

Upgrade to version 7.0.27 and see if that helps. 7.0.25 had some scheduling 'bugs' in it, not 'bugs' exactly, but 'assumed' that everything was being switched to the new way NOW and it didn't happen so was problematic!

This issue still exists for me. Using BOINC 7.0.28... Not using any extra files (app_info.xml, etc.) myself, and while Moo will run regular 1.03 WUs, it still won't run ati14 WUs, but it downloads them. Used to run them just fine under BOINC 6. My credit per day for Moo has basically dropped to insignificance ever since I started using 7.

This issue still exists for me. Using BOINC 7.0.28... Not using any extra files (app_info.xml, etc.) myself, and while Moo will run regular 1.03 WUs, it still won't run ati14 WUs, but it downloads them. Used to run them just fine under BOINC 6. My credit per day for Moo has basically dropped to insignificance ever since I started using 7.

I don't know what the problem is on your end but it seems to work fine on my end, here is a link to one of my completed and validated units:
http://moowrap.net/result.php?resultid=12760278

If you scroll to the start of the details you will see that pc is running Boinc version 7.0.25 and I got Credit 6,144.00 using Application version Distributed.net Client v1.03 (ati14).

ps I went back and looked and you are using a more up to date gpu driver than I am, I am using version 11.9 or driver: 1.4.1546, while you are using driver: 1.4.1720. This could be the problem, do you game? If not try rolling back your driver and see if that fixes the problem. Right now I have 9 pc's with their gpu's attached here and all are using at least Boinc version 7.0.25 and are crunching just fine.

I don't know what the problem is on your end but it seems to work fine on my end, here is a link to one of my completed and validated units:
http://moowrap.net/result.php?resultid=12760278

If you scroll to the start of the details you will see that pc is running Boinc version 7.0.25 and I got Credit 6,144.00 using Application version Distributed.net Client v1.03 (ati14).

ps I went back and looked and you are using a more up to date gpu driver than I am, I am using version 11.9 or driver: 1.4.1546, while you are using driver: 1.4.1720. This could be the problem, do you game? If not try rolling back your driver and see if that fixes the problem. Right now I have 9 pc's with their gpu's attached here and all are using at least Boinc version 7.0.25 and are crunching just fine.

I do game. I suppose it's possible that it could be the driver - 1.4.1720 must be the same as 12.4. I'd rather not downgrade my video drivers just for number crunching when I'm not at the machine, though.

I don't know what the problem is on your end but it seems to work fine on my end, here is a link to one of my completed and validated units:
http://moowrap.net/result.php?resultid=12760278

If you scroll to the start of the details you will see that pc is running Boinc version 7.0.25 and I got Credit 6,144.00 using Application version Distributed.net Client v1.03 (ati14).

ps I went back and looked and you are using a more up to date gpu driver than I am, I am using version 11.9 or driver: 1.4.1546, while you are using driver: 1.4.1720. This could be the problem, do you game? If not try rolling back your driver and see if that fixes the problem. Right now I have 9 pc's with their gpu's attached here and all are using at least Boinc version 7.0.25 and are crunching just fine.

I do game. I suppose it's possible that it could be the driver - 1.4.1720 must be the same as 12.4. I'd rather not downgrade my video drivers just for number crunching when I'm not at the machine, though.

I agree with you!

Hello,

Hmm, if other projects are working and we are not then it must be us setting the "use all GPUs on the host for crunching one WU" switch. New BOINC Client must have some critical bugs regarding that setting where they think we never have enough resources to crunch. :(

I'll try to find some time to upgrade my host to latest stable BOINC Client and see what happens.

-w

That would be most appreciated. I have 2 machines that I am currently using for Moo that use 6.12.34 - it is the ONLY way that I have been able to run Moo at all.

I have tried going back to earlier versions of Catalyst (11.6) and several in between all the way to current - 12.6 - with no luck. Computer 12650 runs Vista 64-bit and computer 15082 runs Win7 64-bit.

With 7.0.25 or 28, both will download 6-12 work units but refuse to run. I have not tried waiting on them to "time-out" so I don't know whether the scheduler will force them to run high priority.

---

That would be most appreciated. I have 2 machines that I am currently using for Moo that use 6.12.34 - it is the ONLY way that I have been able to run Moo at all.

I have tried going back to earlier versions of Catalyst (11.6) and several in between all the way to current - 12.6 - with no luck. Computer 12650 runs Vista 64-bit and computer 15082 runs Win7 64-bit.

With 7.0.25 or 28, both will download 6-12 work units but refuse to run. I have not tried waiting on them to "time-out" so I don't know whether the scheduler will force them to run high priority.

I did the same thing and even set my other gpu projects to NNT and it made no difference. I even detached from my other projects and it didn't help, Moo just wouldn't run! I even did a 'refresh' no help, I did a detach and reattach, no difference. Moo would download units but REFUSE to run them!! I know run other projects on those pc's!

Hello,

Hmm, if other projects are working and we are not then it must be us setting the "use all GPUs on the host for crunching one WU" switch. New BOINC Client must have some critical bugs regarding that setting where they think we never have enough resources to crunch. :(

I'll try to find some time to upgrade my host to latest stable BOINC Client and see what happens.

-w

Teemu, if you could give us a switch that will allow BOINC to think that it is going to use only 1 GPU it would really help. The dnet app still uses all GPUs but that way we can run it along with something like POEM that has low GPU usage and Moo will will use the rest of the GPU that's not utilized. I have one box with 2 5850 cards that does exactly that and it works well. Moo uses 1.5 GPUs and POEM uses .5. For some reason that box downloads Moo WUs with the settings of: 0.02C + 1 ATI GPU. This allows the above behavior to work. My other boxes always DL Moo WUs with 0.02C + 2 ATI GPU, don't know why there's a differences. But if we could force the 0.02C + 1 ATI GPU setting it would be great.

Edit: Never mind, finally figured out how to do it with an app_info. One box had the app_info and the others didn't, duh :)

Hello,

After updating my two workstations to 7.0.28 (latest stable available at the moment) it does indeed seem there's a BOINC Client bug when a task requires more than one GPU resource to run.

My first workstation, that currently only has 1 ATI, is still running tasks just fine. My other workstation, that currently has 2 nVidia cards just downloads them but BOINC Client never runs them. Here's the relevant log entries:

NVIDIA GPU 0: GeForce GTX 285 (driver version 275.33, CUDA version 4.0, compute capability 1.3, 967MB, 918MB available, 1063 GFLOPS peak)
NVIDIA GPU 1: GeForce GTX 285 (driver version 275.33, CUDA version 4.0, compute capability 1.3, 967MB, 919MB available, 1063 GFLOPS peak)
OpenCL: NVIDIA GPU 0: GeForce GTX 285 (driver version 275.33, device version OpenCL 1.0 CUDA, 967MB, 918MB available)
OpenCL: NVIDIA GPU 1: GeForce GTX 285 (driver version 275.33, device version OpenCL 1.0 CUDA, 967MB, 919MB available)
...
[cpu_sched_debug] insufficient NVIDIA for dnetc_r72_1342286579_4_247_0

And the task has 0.2C + 2NV so there's clearly enough nVidia cards to go around. There's not much I can do to fix that (other than develop one task per GPU support), so it's up to the BOINC Client devs to fix their scheduling bug. Has anybody reported this to them yet?

-w

Edit: Never mind, finally figured out how to do it with an app_info. One box had the app_info and the others didn't, duh :)

Indeed, but I could still add an advanced scheduler option that makes the scheduler to only ask for one resource even though the tasks will use all GPUs in the end. You just can't blame me then if you systems overheat or crawl to a stop due to overcommitting GPU resources. ;)

This would probably be easier than to make the full support for running one task per GPU but this would be first step anyway. (Such scheduler modification is needed anyway eventually.)

-w

Running 7.0.28 in Vista with 1 HD4850 (i7 920 host) and it's going just fine.