Problem with RTX2080

Forums › Cruncher's Corner

Dougga
Dougga
Joined: 27 Nov 06
Posts: 27
Credit: 24844941
RAC: 0
12 Dec 2018 10:11:45 UTC
Topic 217326
(moderation:
    )

    Hi,

     

    I just received a new monitor an Acer X34P with G-Sync to go with my Nvidia RTX2080.

    The resolution is set to 3440 x 1440 (considered Native)

     

    Why I  mention this is this rig was working fairly well with both the integrated Intel GPU and the Nvidia.

    Now, with the above monitor plugged in, I get either locked Nvidia computation (meaning no progress) or 100% computational errors with my monitor freezing for a second or two every 20s which seems to be associated with throwing comp. errors in Boinc.

    Currently, I simply get loads of errors which is blocking most of my boinc "productivity".

     

    Any help would be appreciated.

    Thanks

     

    Gary Roberts
    Gary Roberts
    Moderator
    Joined: 9 Feb 05
    Posts: 5872
    Credit: 117010004342
    RAC: 36490856

    Please go to the Problems

    12 Dec 2018 10:19:38 UTC
    Message 168207
    (moderation:
      )

      Please go to the Problems board and read the several current threads that deal with problems with Turing series GPUs.

      There is also a *lot* of information in the Turing thread on this board.

       

      Cheers,
      Gary.

      archae86
      archae86
      Joined: 6 Dec 05
      Posts: 3157
      Credit: 7205854931
      RAC: 928988

      Thanks for the report. I

      12 Dec 2018 14:02:52 UTC
      Message 168212
      (moderation:
        )

        Thanks for the report.

        I reviewed the stderr for this task today and consider that failure to be in the standard behavior pattern for Windows hosts running 2001L series Einstein GRP tasks.

        Fast failure with elapsed times a little over twenty seconds, exit code 28, and this section near the end of stderr all seem typical, sadly.

        % Filling array of photon pairs
ERROR: /home/bema/fermilat/src/bridge_fft_clfft.c:948: clFinish failed. status=-36
ERROR: opencl_ts_2_phase_diff_sorted() returned with error 240867616
01:43:18 (7208): [CRITICAL]: ERROR: MAIN() returned with error '-36'
FPU status flags:  PRECISION
01:43:30 (7208): [normal]: done. calling boinc_finish(28).

        Thank you for the report.  I currently have a list of eleven Einstein hosts displaying this problem running two series of GRP WUs on Turing cards.

        This particular portion of your trouble has nothing to do with your monitor, I believe.

         

         

        Dougga
        Dougga
        Joined: 27 Nov 06
        Posts: 27
        Credit: 24844941
        RAC: 0

        Thanks for the review. I

        12 Dec 2018 21:32:45 UTC
        Message 168221
        (moderation:
          )

          Thanks for the review.

          I turned on the Nvidia GPU and it quickly threw errors on all the available work units and I've reached the limit for today.  The Intel CPU is now working correctly on its workunits but for some reason I only get a tiny amount of credit for these.  I assume I'll throw errors until this Turing issue is resolved.

           

          Thanks again!

          Keith Myers
          Keith Myers
          Joined: 11 Feb 11
          Posts: 4957
          Credit: 18619708143
          RAC: 5619985

          I have other cards in my

          12 Dec 2018 21:46:55 UTC
          Message 168222
          (moderation:
            )

            I have other cards in my host.  My GTX 1080Ti and 1070's still crunch for Einstein.  You can exclude the RTX 2080 from BOINC using it for the Einstein project by adding this to your cc_config.xml in the main BOINC directory. Put it in the <options> section.  Re-read the config files and you should see the gpu being excluded from Einstein.

            <exclude_gpu>

            <url>http://einstein.phys.uwm.edu/</url>
            <device_num>0</device_num>

            </exclude_gpu>

            You will have to verify which device BOINC thinks is the device number<N>.  Probably the 2080 will be identified as Device 0 since it is the most powerful and BOINC generally lists the most capable device first.  The Intel built-in gpu probably will be listed as Device 1.  Look at the startup of the BOINC Event Log to see how your devices are identified to be sure.

             

            Gandolph1
            Gandolph1
            Joined: 20 Feb 05
            Posts: 180
            Credit: 389632201
            RAC: 710

            FYI - I have escalated this

            16 Dec 2018 19:27:27 UTC
            Message 168292
            (moderation:
              )

              FYI - I have escalated this problem with Nvidia directly;

               


              Hello,

              Your case is being escalated to our Level 2 Tech Support group.  The Level 2 agents will review the case notes and may attempt to recreate your issue or find a workaround solution if possible.  As this process may take some time we ask that you be patient and a Level 2 tech will contact you as soon they can to help resolve your issue.

              Best Regards,

              NVIDIA Customer Care

               


              Customer By CSS Web (xxxx xxxxx) (12/16/2018 07:53 AM)


              I have attached the application dump files the system generates when trying to run the Einstein@home task. It generated 9 of these in the very short space of time I allowed it to run. If you need the other 5 let me know.

               

              Comment viewing options

              Select your preferred way to display the comments and click "Save settings" to activate your changes.

              Forums › Cruncher's Corner