ByeBye Wingman
log in

Advanced search

Message boards : News : ByeBye Wingman

1 · 2 · 3 · Next
Author Message
Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 1
Message 16284 - Posted: 19 Apr 2013, 16:50:12 UTC

A new application has been added named "solo_collatz". This application has a quorum of 1 so credit will be granted as soon as the result is validated. Rather than wait for a second computer to return the same result, the solo application uses the CPU to double check the GPU results eliminating the need for a wingman. Multiple instances of the solo CUDA and OpenCL applications can be run on a GPU at a time via the use of an app_config.xml file if GPU utilization is lower than with previous Collatz applications.

If you wish to run both GPU and CPU applications, it is suggested that you uncheck both "collatz" and "mini_collatz", check "solo_collatz" and also check "If no work for selected applications is available, accept work from other applications?" That will enable you to get mini_collatz work for you CPU and solo_collatz work for your GPU.

Dirk Broer
Send message
Joined: 20 Aug 10
Posts: 33
Credit: 176,079,185
RAC: 556,561
Message 16301 - Posted: 20 Apr 2013, 14:28:11 UTC - in response to Message 16284.

Would it be possible to post a good working app_config.xml for Collatz? Presently my HD 6670 takes more than 6 hours for a mini_collatz, under Ati Stream that was 14 minutes.....
____________

Profile L@MiR
Send message
Joined: 12 Jul 09
Posts: 1
Credit: 46,230,522
RAC: 0
Message 16306 - Posted: 21 Apr 2013, 9:44:09 UTC
Last modified: 21 Apr 2013, 9:45:02 UTC

the solo application uses the CPU to double check the GPU results eliminating the need for a wingman.


That will enable you to get mini_collatz work for you CPU and solo_collatz work for your GPU.

??? I'm confused!

bones cruncher
Send message
Joined: 14 Jul 12
Posts: 3
Credit: 19,768,680
RAC: 0
Message 16310 - Posted: 21 Apr 2013, 15:30:17 UTC

All of the solo units are immediately failing on my computer. Not sure what I need to do to fix.

Profile Raubhautz*
Avatar
Send message
Joined: 22 Sep 12
Posts: 16
Credit: 3,373,765
RAC: 0
Message 16314 - Posted: 21 Apr 2013, 22:04:13 UTC - in response to Message 16284.

A new application has been added named "solo_collatz". This application has a quorum of 1 so credit will be granted as soon as the result is validated. Rather than wait for a second computer to return the same result, the solo application uses the CPU to double check the GPU results eliminating the need for a wingman. Multiple instances of the solo CUDA and OpenCL applications can be run on a GPU at a time via the use of an app_config.xml file if GPU utilization is lower than with previous Collatz applications.

If you wish to run both GPU and CPU applications, it is suggested that you uncheck both "collatz" and "mini_collatz", check "solo_collatz" and also check "If no work for selected applications is available, accept work from other applications?" That will enable you to get mini_collatz work for you CPU and solo_collatz work for your GPU.


I did as this post suggests, and all I get are 4.2 mini's and collatz' wu. I have config for 5.0 and get none; message claims I don't have config for app_info (which I then guesstimated), and it still fails. I even tried to reset the project to 'attract' the solo files, but no dice.

So, 2 questions:

1. Why am I not getting cuda5.0 wu?
2. Where is this config for solo wu?

Thank you in advance.
____________

Profile mikey
Avatar
Send message
Joined: 11 Aug 09
Posts: 3242
Credit: 1,691,830,500
RAC: 5,656,435
Message 16316 - Posted: 21 Apr 2013, 22:33:48 UTC - in response to Message 16314.
Last modified: 21 Apr 2013, 22:36:11 UTC

A new application has been added named "solo_collatz". This application has a quorum of 1 so credit will be granted as soon as the result is validated. Rather than wait for a second computer to return the same result, the solo application uses the CPU to double check the GPU results eliminating the need for a wingman. Multiple instances of the solo CUDA and OpenCL applications can be run on a GPU at a time via the use of an app_config.xml file if GPU utilization is lower than with previous Collatz applications.

If you wish to run both GPU and CPU applications, it is suggested that you uncheck both "collatz" and "mini_collatz", check "solo_collatz" and also check "If no work for selected applications is available, accept work from other applications?" That will enable you to get mini_collatz work for you CPU and solo_collatz work for your GPU.


I did as this post suggests, and all I get are 4.2 mini's and collatz' wu. I have config for 5.0 and get none; message claims I don't have config for app_info (which I then guesstimated), and it still fails. I even tried to reset the project to 'attract' the solo files, but no dice.

So, 2 questions:

1. Why am I not getting cuda5.0 wu?
2. Where is this config for solo wu?

Thank you in advance.


When I see my computers I see this on one of them:
INTEL Intel(R) HD Graphics 4000 (1624MB) OpenCL: 1.02

I do NOT see that on any of yours.
Oops never mind I found it "OpenCL: 1.01".

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 1
Message 16322 - Posted: 22 Apr 2013, 0:26:57 UTC - in response to Message 16301.

The values are GPU specific, so I can't tell you the proper values unless you want to buy me an exact duplicate of you GPU so I can do the trail and error to figure it out.
Rules: items_per_kernel + kernels_per_reduction < 31
If kernels per reduction is too high, it starts using CPU
If threads is too high, it will crash. 5 is lowest value (2^5 = 32). 9 is max (2^9 = 256) on HD 6970.

<exe_name>.config file:
verbose=1
items_per_kernel=20
kernels_per_reduction=9
threads=8
sleepMS=1



app_config.xml (easier way to run multiple WUs at once):
<app_config>
<app>
<name>solo_collatz</name>
<max_concurrent>3</max_concurrent>
<gpu_versions>
<gpu_usage>0.33</gpu_usage>
<cpu_usage>.01</cpu_usage>
</gpu_version>
</app>
</app_config>

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 1
Message 16323 - Posted: 22 Apr 2013, 0:28:46 UTC - in response to Message 16310.

All of the solo units are immediately failing on my computer. Not sure what I need to do to fix.



Use the 12.7 Catalyst driver. The 12.10 may work as well. 13.1 or later fail. AMD says they __should__ have it fixed in version 13.4. Until then, there's nothing I can do.

bones cruncher
Send message
Joined: 14 Jul 12
Posts: 3
Credit: 19,768,680
RAC: 0
Message 16325 - Posted: 22 Apr 2013, 2:59:16 UTC - in response to Message 16323.

All of the solo units are immediately failing on my computer. Not sure what I need to do to fix.



Use the 12.7 Catalyst driver. The 12.10 may work as well. 13.1 or later fail. AMD says they __should__ have it fixed in version 13.4. Until then, there's nothing I can do.


Unfortunately, I went to Dell's web page, and the older driver is not available. Can't use drivers direct from AMD. They will not work with my Dell. I'll just have to shut off the solo. The regular collatz works fine.

Kombizahl
Send message
Joined: 29 Sep 09
Posts: 12
Credit: 158,620,174
RAC: 204
Message 16326 - Posted: 22 Apr 2013, 10:38:09 UTC

Here dito,
all solo tasks became compute errors.
Greetings

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 1
Message 16329 - Posted: 22 Apr 2013, 16:11:02 UTC - in response to Message 16326.

Here dito,
all solo tasks became compute errors.
Greetings


It looks like it is failing as soon as BOINC starts the application as it isn't even logging the name of the app which is the first thing it does.

Are there any clues in ay of the Windows event logs as to the cause of the error?
What happens if you run the app in standalone mode? (e.g. Copy one of the workunits to "in.txt" and run the executable outside of BOINC. Then post the contents of stderr.txt.)

Profile Slicker
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 11 Jun 09
Posts: 2525
Credit: 740,580,099
RAC: 1
Message 16330 - Posted: 22 Apr 2013, 16:25:14 UTC - in response to Message 16325.

All of the solo units are immediately failing on my computer. Not sure what I need to do to fix.



Use the 12.7 Catalyst driver. The 12.10 may work as well. 13.1 or later fail. AMD says they __should__ have it fixed in version 13.4. Until then, there's nothing I can do.


Unfortunately, I went to Dell's web page, and the older driver is not available. Can't use drivers direct from AMD. They will not work with my Dell. I'll just have to shut off the solo. The regular collatz works fine.


Have you tried getting the drivers directly from AMD since that is where Dell gets them from? I'd try version 12.6 first. If that doesn't work, then 12.8 and finally 12.10. I would stay away from 13.1 since it doesn't work for a large number of people.

http://support.amd.com/us/gpudownload/windows/previous/Pages/radeonaiw_vista64.aspx (While it says Vista, they are also the same drivers for Windows 7 x64.)

bones cruncher
Send message
Joined: 14 Jul 12
Posts: 3
Credit: 19,768,680
RAC: 0
Message 16337 - Posted: 22 Apr 2013, 23:19:43 UTC - in response to Message 16330.

All of the solo units are immediately failing on my computer. Not sure what I need to do to fix.



Use the 12.7 Catalyst driver. The 12.10 may work as well. 13.1 or later fail. AMD says they __should__ have it fixed in version 13.4. Until then, there's nothing I can do.


Unfortunately, I went to Dell's web page, and the older driver is not available. Can't use drivers direct from AMD. They will not work with my Dell. I'll just have to shut off the solo. The regular collatz works fine.


Have you tried getting the drivers directly from AMD since that is where Dell gets them from? I'd try version 12.6 first. If that doesn't work, then 12.8 and finally 12.10. I would stay away from 13.1 since it doesn't work for a large number of people.

http://support.amd.com/us/gpudownload/windows/previous/Pages/radeonaiw_vista64.aspx (While it says Vista, they are also the same drivers for Windows 7 x64.)

Dell has their own modified drivers. The last time I tried to get drivers direct from AMD, I ended up messing things up and had to have Dell support fix what I did. I was informed by the Dell technician that I must get my drivers directly from them, not AMD.

Profile arkayn
Volunteer tester
Avatar
Send message
Joined: 30 Aug 09
Posts: 219
Credit: 676,877,192
RAC: 21,486
Message 16338 - Posted: 22 Apr 2013, 23:24:28 UTC - in response to Message 16337.

All of the solo units are immediately failing on my computer. Not sure what I need to do to fix.



Use the 12.7 Catalyst driver. The 12.10 may work as well. 13.1 or later fail. AMD says they __should__ have it fixed in version 13.4. Until then, there's nothing I can do.


Unfortunately, I went to Dell's web page, and the older driver is not available. Can't use drivers direct from AMD. They will not work with my Dell. I'll just have to shut off the solo. The regular collatz works fine.


Have you tried getting the drivers directly from AMD since that is where Dell gets them from? I'd try version 12.6 first. If that doesn't work, then 12.8 and finally 12.10. I would stay away from 13.1 since it doesn't work for a large number of people.

http://support.amd.com/us/gpudownload/windows/previous/Pages/radeonaiw_vista64.aspx (While it says Vista, they are also the same drivers for Windows 7 x64.)

Dell has their own modified drivers. The last time I tried to get drivers direct from AMD, I ended up messing things up and had to have Dell support fix what I did. I was informed by the Dell technician that I must get my drivers directly from them, not AMD.


I think I know what I would tell that Dell tech {NSFW}.

There should not be modifying drivers for third party graphics cards.

____________

Profile mikey
Avatar
Send message
Joined: 11 Aug 09
Posts: 3242
Credit: 1,691,830,500
RAC: 5,656,435
Message 16344 - Posted: 23 Apr 2013, 11:20:11 UTC - in response to Message 16337.

All of the solo units are immediately failing on my computer. Not sure what I need to do to fix.



Use the 12.7 Catalyst driver. The 12.10 may work as well. 13.1 or later fail. AMD says they __should__ have it fixed in version 13.4. Until then, there's nothing I can do.


Unfortunately, I went to Dell's web page, and the older driver is not available. Can't use drivers direct from AMD. They will not work with my Dell. I'll just have to shut off the solo. The regular collatz works fine.


Have you tried getting the drivers directly from AMD since that is where Dell gets them from? I'd try version 12.6 first. If that doesn't work, then 12.8 and finally 12.10. I would stay away from 13.1 since it doesn't work for a large number of people.

http://support.amd.com/us/gpudownload/windows/previous/Pages/radeonaiw_vista64.aspx (While it says Vista, they are also the same drivers for Windows 7 x64.)

Dell has their own modified drivers. The last time I tried to get drivers direct from AMD, I ended up messing things up and had to have Dell support fix what I did. I was informed by the Dell technician that I must get my drivers directly from them, not AMD.


I have a Dell laptop and downloaded the 12.6 drivers and after loading it said 'no adapters found', meaning it didn't work. Still no solo units though.

Profile Mark W. Patton
Send message
Joined: 20 Apr 10
Posts: 20
Credit: 788,413,955
RAC: 4,802,710
Message 16363 - Posted: 25 Apr 2013, 12:30:44 UTC

I am concerned about this change. Milkyway changed their apps a while back and rendered two of my machines useless on their project. It is beginning to look like volunteered computers are irrelevant. I hope the same thing does not happen here.

Profile Raubhautz*
Avatar
Send message
Joined: 22 Sep 12
Posts: 16
Credit: 3,373,765
RAC: 0
Message 16368 - Posted: 26 Apr 2013, 7:28:50 UTC - in response to Message 16363.

I am concerned about this change. Milkyway changed their apps a while back and rendered two of my machines useless on their project. It is beginning to look like volunteered computers are irrelevant. I hope the same thing does not happen here.



Not sure about MW, but I can no longer use Collatz on my machines... Which is too bad. It was running the 23's fine, but now, just errors. I had to move to POEM to keep my gpu's busy.
____________

Profile Mark W. Patton
Send message
Joined: 20 Apr 10
Posts: 20
Credit: 788,413,955
RAC: 4,802,710
Message 16370 - Posted: 26 Apr 2013, 12:47:58 UTC - in response to Message 16368.
Last modified: 26 Apr 2013, 13:01:05 UTC

That is what I intend to do. As much as I hate leaving projects that I have devoted so much time and resources to, I am just going to run the ones that work. It is a shame. I have to wonder if they were not getting enough data processed. You know, I currently process 25 threads of data 24/7. I am thinking I will throw all my resources at one project-whichever one works. :)

Profile mikey
Avatar
Send message
Joined: 11 Aug 09
Posts: 3242
Credit: 1,691,830,500
RAC: 5,656,435
Message 16376 - Posted: 27 Apr 2013, 11:14:55 UTC - in response to Message 16368.

I am concerned about this change. Milkyway changed their apps a while back and rendered two of my machines useless on their project. It is beginning to look like volunteered computers are irrelevant. I hope the same thing does not happen here.



Not sure about MW, but I can no longer use Collatz on my machines... Which is too bad. It was running the 23's fine, but now, just errors. I had to move to POEM to keep my gpu's busy.


Umm that might not work, Poem is regulating the amount of gpu units per day to the number they can actually review, meaning it is not nearly enough to keep everyones caches full and people often run out of them.

Profile mikey
Avatar
Send message
Joined: 11 Aug 09
Posts: 3242
Credit: 1,691,830,500
RAC: 5,656,435
Message 16377 - Posted: 27 Apr 2013, 11:22:34 UTC - in response to Message 16370.

That is what I intend to do. As much as I hate leaving projects that I have devoted so much time and resources to, I am just going to run the ones that work. It is a shame. I have to wonder if they were not getting enough data processed. You know, I currently process 25 threads of data 24/7. I am thinking I will throw all my resources at one project-whichever one works. :)


I do that too, except I tend to split my cpu's and gpu's to different projects, each trying for a different goal. I am a goal oriented guy so it is easy for me to settle in for the long haul and finish what I started. Right now my gpu's are at 3 different places just because I have different goals at each. My cpu's too are at 2 different places, one primary and another set at 0% for those times when the first runs out of work. I have both AMD and Nvidia gpu's so splitting them between projects only makes sense as each can be optimized in their crunching. I also have an AMD 7970 gpu and it is crunching at its project because it works there and not at the place I really want it to be at.

1 · 2 · 3 · Next
Post to thread

Message boards : News : ByeBye Wingman


Main page · Your account · Message boards


Copyright © 2018 Jon Sonntag; All rights reserved.