all 14au18ac GPU tasks on one system running 3 seconds?

Message boards : Number crunching : all 14au18ac GPU tasks on one system running 3 seconds?
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · Next

AuthorMessage
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 9640
Credit: 886,901,875
RAC: 1,719,302
United States
Message 1950615 - Posted: 18 Aug 2018, 22:42:08 UTC

Hi Petri, then color me very CONFUSED. I downloaded the file from your link. It is named setiV0.97.linux_x64_10x0. So that is what I changed the appname to setiV0.97.linux_x64_10x0 in the file info and app version defines.

I look at your app_info example and see no sign of the 0.97 app name setiV0.97.linux_x64_10x0. I see instead setiathome_x41zc_x86_64-pc-linux-gnu_cuda65_v8.

So is the real name of the 0.97 app setiathome_x41zc_x86_64-pc-linux-gnu_cuda65_v8?

No wonder I dumped all the work if the name definition is not the name of the real executable name.

Also should I remove the libcudart.so.9.0 and libcufft.so.9.0 file defines for the zi3v app? Or just leave them in there and they will cause no harm. I expect they have to be removed. Also I see the -nobs still in there. I though someone said that is no longer effective for the static linked apps?

More clarification please.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1950615 · Report as offensive
Profile petri33
Volunteer tester

Send message
Joined: 6 Jun 02
Posts: 1654
Credit: 534,647,234
RAC: 471,523
Finland
Message 1950619 - Posted: 18 Aug 2018, 22:55:16 UTC - in response to Message 1950615.  
Last modified: 18 Aug 2018, 23:02:26 UTC

Hi Petri, then color me very CONFUSED. I downloaded the file from your link. It is named setiV0.97.linux_x64_10x0. So that is what I changed the appname to setiV0.97.linux_x64_10x0 in the file info and app version defines.

I look at your app_info example and see no sign of the 0.97 app name setiV0.97.linux_x64_10x0. I see instead setiathome_x41zc_x86_64-pc-linux-gnu_cuda65_v8.

So is the real name of the 0.97 app setiathome_x41zc_x86_64-pc-linux-gnu_cuda65_v8?

No wonder I dumped all the work if the name definition is not the name of the real executable name.

Also should I remove the libcudart.so.9.0 and libcufft.so.9.0 file defines for the zi3v app? Or just leave them in there and they will cause no harm. I expect they have to be removed. Also I see the -nobs still in there. I though someone said that is no longer effective for the static linked apps?

More clarification please.



1) You changed the names in the app_config.xml. Good. Check for any typos, spaces, ... just in case.
2) Did the unzip set the file executable bit on? You can do that many ways. I use: chmod ugo+x setiV0.97.linux_x64_10x0
3) When the executable file got deleted at boinc restart I suppose you copied/unzippped it back.
4) The executable can have any name. I copy mine over the years old setiathome_x41zc_x86_64-pc-linux-gnu_cuda65_v8 named file in the boinc/seti folder so I do not have to change app_info or anything. Just suspend. copy. resume. Most people, however, modify app_info.xml to match the executable.
5) You can delete the old lib files from boinc/seti folder. They can be left there as well.
6) -nobs speeds things up if you have spare cores. It makes the process (CPU) to wait in a spin loop for the GPU to respond. Without -nobs the process waits an event in 'sleeping' state and the CPU is free to run other jobs meanwhile.
To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones
ID: 1950619 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 9640
Credit: 886,901,875
RAC: 1,719,302
United States
Message 1950623 - Posted: 18 Aug 2018, 23:19:02 UTC - in response to Message 1950619.  

1) The only change to app_config.xml was to remove the -nobs command line parameter as I was told it is not necessary on the static linked apps because they already are memory efficient with much less memory calls compared to zi3v.

2) Checked the execute bit on the new executable was set after unpacking. It was.

3) Every time the new executable got deleted, I copied it back from its unpack location in /Downloads.

4) I think I figured out your app_name in your app_info example after a while since I finally realized the app_name is the name of the stock Linux app. You just copied over the new executable onto the existing app defined in app_info after your last comment in your first reply.

5) OK, I will leave the -nobs in place as I have plenty of cores to feed the gpus.

6) This is the app_info.xml I tried to run and it ghosted all the gpu cache. I removed the file references to libcudart.so.9.0 and libcufft.so.9.0 file defines. Is that what ghosted my tasks?

<app_info>
  <app>
     <name>setiathome_v8</name>
  </app>
    <file_info>
      <name>setiV0.97.linux_x64_10x0</name>
      <executable/>
    </file_info>
    <file_info>
      <name></name>
    </file_info>
    <file_info>
      <name></name>
    </file_info>
    <app_version>
      <app_name>setiathome_v8</app_name>
      <platform>x86_64-pc-linux-gnu</platform>
      <version_num>801</version_num>
      <plan_class>cuda90</plan_class>
      <cmdline></cmdline>
      <coproc>
        <type>NVIDIA</type>
        <count>1</count>
      </coproc>
      <avg_ncpus>0.1</avg_ncpus>
      <max_ncpus>0.1</max_ncpus>
      <file_ref>
         <file_name>setiV0.97.linux_x64_10x0</file_name>
          <main_program/>
      </file_ref>
      <file_ref>
         <file_name></file_name>
      </file_ref>
      <file_ref>
         <file_name></file_name>
      </file_ref>
    </app_version>
  <app>
     <name>astropulse_v7</name>
  </app>
     <file_info>
       <name>astropulse_7.08_x86_64-pc-linux-gnu__opencl_nvidia_100</name>
        <executable/>
     </file_info>
     <file_info>
       <name>AstroPulse_Kernels_r2751.cl</name>
     </file_info>
     <file_info>
       <name>ap_cmdline_7.08_x86_64-pc-linux-gnu__opencl_nvidia_100.txt</name>
     </file_info>
    <app_version>
      <app_name>astropulse_v7</app_name>
      <platform>x86_64-pc-linux-gnu</platform>
      <version_num>708</version_num>
      <plan_class>opencl_nvidia_100</plan_class>
      <coproc>
        <type>NVIDIA</type>
        <count>1</count>
      </coproc>
      <avg_ncpus>0.1</avg_ncpus>
      <max_ncpus>0.1</max_ncpus>
      <file_ref>
         <file_name>astropulse_7.08_x86_64-pc-linux-gnu__opencl_nvidia_100</file_name>
          <main_program/>
      </file_ref>
      <file_ref>
         <file_name>AstroPulse_Kernels_r2751.cl</file_name>
      </file_ref>
      <file_ref>
         <file_name>ap_cmdline_7.08_x86_64-pc-linux-gnu__opencl_nvidia_100.txt</file_name>
         <open_name>ap_cmdline.txt</open_name>
      </file_ref>
    </app_version>
   <app>
      <name>setiathome_v8</name>
   </app>
      <file_info>
         <name>MBv8_8.22r3711_sse41_x86_64-pc-linux-gnu</name>
         <executable/>
      </file_info>
     <app_version>
     <app_name>setiathome_v8</app_name>
     <platform>x86_64-pc-linux-gnu</platform>
     <version_num>800</version_num>   
      <file_ref>
        <file_name>MBv8_8.22r3711_sse41_x86_64-pc-linux-gnu</file_name>
        <main_program/>
      </file_ref>
    </app_version>
   <app>
      <name>astropulse_v7</name>
   </app>
     <file_info>
       <name>ap_7.05r2728_sse3_linux64</name>
        <executable/>
     </file_info>
    <app_version>
       <app_name>astropulse_v7</app_name>
       <version_num>704</version_num>
       <platform>x86_64-pc-linux-gnu</platform>
       <plan_class></plan_class>
       <file_ref>
         <file_name>ap_7.05r2728_sse3_linux64</file_name>
          <main_program/>
       </file_ref>
    </app_version>
</app_info>


7 ) See some new AP and MB cpu applications listed in your example. Should we expect those at some time at Lunatics or Crunchers Anonymous?
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1950623 · Report as offensive
Profile petri33
Volunteer tester

Send message
Joined: 6 Jun 02
Posts: 1654
Credit: 534,647,234
RAC: 471,523
Finland
Message 1950625 - Posted: 18 Aug 2018, 23:33:18 UTC - in response to Message 1950623.  

in the app_info there are 'empty' parts
    <file_info>
      <name></name>
    </file_info>
    <file_info>
      <name></name>
    </file_info>

I do not know, but the may confuse boinc..
... and to the 7) the names may be old or something. Leftovers from trying to build a linux executable years ago. Definitely nothing new.

I hope you get your system to work.
Could someone else help, since it is bed time in this side of the world (2:30 AM).

Petri
To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones
ID: 1950625 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 9640
Credit: 886,901,875
RAC: 1,719,302
United States
Message 1950627 - Posted: 18 Aug 2018, 23:38:22 UTC - in response to Message 1950625.  

The empty parts are my removal of the dynamic link cuda90 libraries defines.

I'm recovering my ghosts and then I will work the existing cache down to minimum and then try the simple copy the new 0.97 executable over the existing zi3v executable so that app_info is still looking for the same app_name defines. Then I will try again with the 0.97 and if issues at least I will ghost as few as possible tasks.

Thanks for the explanations. Go to bed.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1950627 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 4807
Credit: 552,190,936
RAC: 1,264,078
United States
Message 1950634 - Posted: 18 Aug 2018, 23:59:48 UTC

Well, I've left my app_info.xml unchanged from using the older Apps. I do change the library numbers to match the App I'm using though. I really don't think the numbers matter on the new Apps, as long as BOINC can find the named files. I suppose empty fields may be unwise. So, say I go from using the old CUDA9 version to the current version, all I do is change the App names and library numbers so it looks as though;
  <app>
    <name>setiathome_v8</name>
  </app>
	<file_info>
	   <name>setiathome_x41p_V0.97_x86_64-pc-linux-gnu_cuda91</name>
	   <executable/>
	</file_info>
	<file_info>
	   <name>libcudart.so.9.1</name>
	   <executable/>
	</file_info>
	<file_info>
	   <name>libcufft.so.9.1</name>
	   <executable/>
	</file_info>
	<app_version>
          <app_name>setiathome_v8</app_name>
	  <platform>x86_64-pc-linux-gnu</platform>
          <version_num>801</version_num>
	  <plan_class>cuda90</plan_class>
          <cmdline>-nobs</cmdline>
	    <avg_ncpus>0.1</avg_ncpus>
	    <max_ncpus>0.1</max_ncpus>
	     <coproc>
	       <type>CUDA</type>
	       <count>1.0</count>
	     </coproc>
	<file_ref>
           <file_name>setiathome_x41p_V0.97_x86_64-pc-linux-gnu_cuda91</file_name>
           <main_program/>
	</file_ref>
	<file_ref>
	   <file_name>libcudart.so.9.1</file_name>
	</file_ref>
	<file_ref>
	   <file_name>libcufft.so.9.1</file_name>
	</file_ref>
     </app_version>

The above is working for Me, and I'm not Ghosting any tasks.
This way if I go back to CUDA 9.0 I don't have to make a lot of changes. Just change the App names and library numbers.
I do have the CUDA 9.1 and 9.2 libraries in the folder though, right next to all the other libraries.
I'll experiment later about rather or not you can remove all the library references, I really don't know.

So far, setiathome_x41p_V0.97_x86_64-pc-linux-gnu_cuda91 seems to be working fine on my Maxwell GPUs.
ID: 1950634 · Report as offensive
Profile petri33
Volunteer tester

Send message
Joined: 6 Jun 02
Posts: 1654
Credit: 534,647,234
RAC: 471,523
Finland
Message 1950636 - Posted: 19 Aug 2018, 0:13:16 UTC
Last modified: 19 Aug 2018, 0:17:05 UTC

This is a Linux compilation for sm_50, sm_52, sm 61 and sm_70. It may run with GTX750. I do not have one to test with. It is statically linked and I guess it will run with suitably recent 39X.xx driver. No dev pack installation required. Try and tell.

It is a bit slower on 10x0 than the previous one because the 750 and 9x0 have less shared memory than 10x0.
multiversion: https://drive.google.com/open?id=1rKaagzCyisp-IYHZrX4hwpfEKPoiKI-h
10x0&Volta: https://drive.google.com/open?id=1pe6-p5zn27tXFvvszyGCzo0OfCkkCqDt

Link to source was posted earlier in this thread.
To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones
ID: 1950636 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 9640
Credit: 886,901,875
RAC: 1,719,302
United States
Message 1950638 - Posted: 19 Aug 2018, 0:32:30 UTC
Last modified: 19 Aug 2018, 0:33:05 UTC

What started all this was your comment about get CUDA9.2 folks as that solves the 0.97 app pulse counts on the overflows that you showed in your benchmark runs. So I got the CUDA9.2 toolkit installed. The only thing I didn't do was let the toolkit install its own graphics drivers since that would have been a backlevel for me since I was already up on 396.61 drivers. I also skipped installation of the CUDA examples.

I grabbed Petri's lastest 0.97 compiled executable from his Google Drive link which has a very different name from your zi3v app names.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1950638 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 4807
Credit: 552,190,936
RAC: 1,264,078
United States
Message 1950639 - Posted: 19 Aug 2018, 0:33:34 UTC - in response to Message 1950636.  

Petri, how about posting your Current CUDA app_info.xml section so we can see what You are using? Similar to what I posted. I believe that's what Keith is looking for.
ID: 1950639 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 4807
Credit: 552,190,936
RAC: 1,264,078
United States
Message 1950640 - Posted: 19 Aug 2018, 0:39:26 UTC - in response to Message 1950638.  

What started all this was your comment about get CUDA9.2 folks as that solves the 0.97 app pulse counts on the overflows that you showed in your benchmark runs. So I got the CUDA9.2 toolkit installed. The only thing I didn't do was let the toolkit install its own graphics drivers since that would have been a backlevel for me since I was already up on 396.61 drivers. I also skipped installation of the CUDA examples.

I grabbed Petri's lastest 0.97 compiled executable from his Google Drive link which has a very different name from your zi3v app names.

All I said was CUDA 9.2 DRIVER. I don't think you will see any mention of the TOOLKIT in my post.
You know, the DRIVER you have been saying NOT to use for the past couple Days.

Personally, it will be a few days before you see any Apps from Me. I'm not into posting Apps within Hours of receiving the Source code. I do believe in testing a while.
So, you still have time to get those CUDA 9.2 DRIVERS installed. I believe that's what was in the first post.
ID: 1950640 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 9640
Credit: 886,901,875
RAC: 1,719,302
United States
Message 1950641 - Posted: 19 Aug 2018, 0:40:49 UTC

He already posted his app_info. His app_info shows the old stock CUDA65 app_name for gpu tasks. That is what he named his new 0.97 app in his app_info so he wouldn't have to make changes to his existing app_info. That is what confused me initially.
So are running with the CUDA91 or CUDA92 Toolkit, TBar?
This is my output from nvcc.
keith@Chomsky:~$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2018 NVIDIA Corporation
Built on Tue_Jun_12_23:07:04_CDT_2018
Cuda compilation tools, release 9.2, V9.2.148

Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1950641 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 9640
Credit: 886,901,875
RAC: 1,719,302
United States
Message 1950642 - Posted: 19 Aug 2018, 0:47:09 UTC - in response to Message 1950640.  

What started all this was your comment about get CUDA9.2 folks as that solves the 0.97 app pulse counts on the overflows that you showed in your benchmark runs. So I got the CUDA9.2 toolkit installed. The only thing I didn't do was let the toolkit install its own graphics drivers since that would have been a backlevel for me since I was already up on 396.61 drivers. I also skipped installation of the CUDA examples.

I grabbed Petri's lastest 0.97 compiled executable from his Google Drive link which has a very different name from your zi3v app names.

All I said was CUDA 9.2 DRIVER. I don't think you will see any mention of the TOOLKIT in my post.
You know, the DRIVER you have been saying NOT to use for the past couple Days.

Personally, it will be a few days before you see any Apps from Me. I'm not into posting Apps within Hours of receiving the Source code. I do believe in testing a while.
So, you still have time to get those CUDA 9.2 DRIVERS installed. I believe that's what was in the first post.

Well that is what has confused me from your post. I never knew that referring to the CUDA 9.2 driver was you calling it that when simply referring to the driver version number. I already was running the 396.51 DRIVER.

I NEVER said not to run the 396.51 driver. I just said it has issues after installation and subsequent use of its nvidia-xserver application and that you MIGHT have issues with changes in xorg.conf.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1950642 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 4807
Credit: 552,190,936
RAC: 1,264,078
United States
Message 1950646 - Posted: 19 Aug 2018, 1:15:22 UTC - in response to Message 1950642.  
Last modified: 19 Aug 2018, 1:19:02 UTC


Keith Myers ; Message 1950359 - Posted: 17 Aug 2018, 19:02:42 UTC
The current short term release 396.51 drivers are pretty buggy to say the least. A lot more than usual. The last good short term release driver was the 396.24 and if you want to be entirely safe, stick with the current long term release 390.77 drivers.
It certainly looks as though you are saying to use 390 drivers there.
Which is why I spent Way too much time asking why you found Bugs in a Driver where I found None.
That post of yours mentions 32 bit drivers, the CUDA Apps are 64 bit, meaning you use the 64 bit drivers. So, I wouldn't think 32 bit troubles have any bearing on people using the SETI CUDA App.

I was asking Petri for his CUDA Section, like mine, something people may understand.
Since His are the only Apps posted, you need to know Which Toolkit He used, not which one I used.
I really don't know which he used, I would guess 9.2.
When I post an App, I will tell you what you need.
ID: 1950646 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 9640
Credit: 886,901,875
RAC: 1,719,302
United States
Message 1950647 - Posted: 19 Aug 2018, 1:33:23 UTC - in response to Message 1950646.  

Lets's just drop this as we are usual talking cross purposes and posts. Semantic definitions are obviously different between us. I was just reporting that I found a half dozen similar posts over on the Ubuntu forums detailing exactly the same issues I have found with the 396.45 and 396.51 drivers. I am not a single case obviously as I found on 5 different systems of others exactly the same issues when installing the latest ppa drivers.

Bully for you that you have not had any issues installing the .run file installation from Nvidia website when I was explicitly saying the drivers from the ppa were troublesome. Not that they didn't install or didn't work, but that they MAY cause configuration issues.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1950647 · Report as offensive
Profile petri33
Volunteer tester

Send message
Joined: 6 Jun 02
Posts: 1654
Credit: 534,647,234
RAC: 471,523
Finland
Message 1950669 - Posted: 19 Aug 2018, 6:35:28 UTC - in response to Message 1950639.  

Petri, how about posting your Current CUDA app_info.xml section so we can see what You are using? Similar to what I posted. I believe that's what Keith is looking for.


That was my app_info.xml and app_config.xml.
I have an ancient executable name there. I always copy my latest over the previous executable file.
I do not change my app_info unless I need to.

That is why I said this is an example app_info.xml. It works, but you have to copy any executable you use over the original cuda 65 bla bla executable.
To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones
ID: 1950669 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 9640
Credit: 886,901,875
RAC: 1,719,302
United States
Message 1950671 - Posted: 19 Aug 2018, 7:04:33 UTC - in response to Message 1950669.  

We got that and understood eventually Petri. Mostly just doing a little light editing of my app_info files to change over to the Pascal or the multiverse app and remove the old cuda library references. Just using the app_names you gave them in the zip files. All looking good so far and VERY impressed by the task speedups.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1950671 · Report as offensive
Profile petri33
Volunteer tester

Send message
Joined: 6 Jun 02
Posts: 1654
Credit: 534,647,234
RAC: 471,523
Finland
Message 1950752 - Posted: 19 Aug 2018, 18:24:17 UTC - in response to Message 1950671.  

We got that and understood eventually Petri. Mostly just doing a little light editing of my app_info files to change over to the Pascal or the multiverse app and remove the old cuda library references. Just using the app_names you gave them in the zip files. All looking good so far and VERY impressed by the task speedups.


Nice to hear. Now let's keep an eye on inconclusives and nvalids.
How hot do your GPUs run? Have you increased fan speed?
I run my fans at 100%
To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones
ID: 1950752 · Report as offensive
Profile Zalster Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 27 May 99
Posts: 5178
Credit: 430,787,934
RAC: 329,844
United States
Message 1950754 - Posted: 19 Aug 2018, 18:34:18 UTC - in response to Message 1950752.  

Nice to hear. Now let's keep an eye on inconclusives and nvalids.
How hot do your GPUs run? Have you increased fan speed?
I run my fans at 100%



I was able to upgrade 1 computer. Unfortunately all the rest are at home and will have to wait.

I'm running hybrids but have a bash to run fans on cards at 100% Temps are 35C-42C across the 4 cards.

Thanks Petri. (thanks Keith for the bash)
ID: 1950754 · Report as offensive
Ian&Steve C.
Avatar

Send message
Joined: 28 Sep 99
Posts: 1718
Credit: 699,628,091
RAC: 2,417,669
United States
Message 1950755 - Posted: 19 Aug 2018, 18:35:56 UTC
Last modified: 19 Aug 2018, 18:37:07 UTC

Keith,

all i did when switching to version 0.96 and 0.97 from zi3v:

1. obviously make sure you have a 396.xx version of nvidia drivers installed
2. don't change your app_info.xml file at all from whatever you had previously running zi3v
3. stop BOINC processing, clicking "suspend" under the projects tab was sufficient for me, but you can close it out completely if you wish.
4. backup your current zi3v app somewhere safe, move it to another directory
5. RENAME the v0.97 app filename to exactly match the filename of the old zi3v app. literally just copy and paste the filename
6. place the renamed v0.97 app in the seti directory where the old zi3v app was previously. (BOINC will think it's the same app because it now has the same filename)
7. Resume your SETI project, or re-launch BOINC.
8. Done.

this process worked for me across 6 different systems. only takes a few minutes to do
Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

ID: 1950755 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 13125
Credit: 148,770,865
RAC: 177,204
United Kingdom
Message 1950756 - Posted: 19 Aug 2018, 18:38:41 UTC - in response to Message 1950669.  

That was my app_info.xml and app_config.xml.
I have an ancient executable name there. I always copy my latest over the previous executable file.
I do not change my app_info unless I need to.

That is why I said this is an example app_info.xml. It works, but you have to copy any executable you use over the original cuda 65 bla bla executable.
If I may say so, I think that's a very silly way of doing things.

An example app_info file is included in the BOINC documentation - and that's where examples should stay.

Working applications - or applications intended to be working applications - should have a matching app_info snippet included in the download: I believe Raistmer has scripted his build process to generate such a resource with every build.

It's reasonable to expect testers to roll their own files, but not end users.
ID: 1950756 · Report as offensive
Previous · 1 · 2 · 3 · Next

Message boards : Number crunching : all 14au18ac GPU tasks on one system running 3 seconds?


 
©2019 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.