No work for GPU?

Forums › Problems and Bug Reports

Richard
Richard
Joined: 29 Apr 10
Posts: 39
Credit: 39269698
RAC: 0
19 Nov 2012 4:21:07 UTC
Topic 196634
(moderation:
    )

    I have a new computer, with an i7, and Nvidia graphics. The graphics card is recognized, but when it requests work it gets 0 tasks for the GPU.

    Also, I occasionally see it stop processing tasks, and get a message in the log that "Suspending computation, requested by operating system" Can someone give me some idea why windows would be requesting computation be suspended. This is happening when nothing is running, and no-one is near the machine.

    Please see the log file below.

    11/18/2012 6:40:08 PM | Einstein@Home | Scheduler request completed: got 0 new tasks
    11/18/2012 6:40:08 PM | Einstein@Home | No work sent
    11/18/2012 6:40:08 PM | Einstein@Home | see scheduler log messages on http://einstein.phys.uwm.edu//host_sched_logs/6110/6110411
    11/18/2012 7:07:23 PM | Einstein@Home | Sending scheduler request: To fetch work.
    11/18/2012 7:07:23 PM | Einstein@Home | Requesting new tasks for NVIDIA
    11/18/2012 7:07:24 PM | Einstein@Home | Scheduler request completed: got 0 new tasks
    11/18/2012 7:07:24 PM | Einstein@Home | No work sent
    11/18/2012 7:07:24 PM | Einstein@Home | see scheduler log messages on http://einstein.phys.uwm.edu//host_sched_logs/6110/6110411
    11/18/2012 7:27:33 PM | Einstein@Home | Sending scheduler request: To fetch work.
    11/18/2012 7:27:33 PM | Einstein@Home | Requesting new tasks for NVIDIA
    11/18/2012 7:27:35 PM | Einstein@Home | Scheduler request completed: got 0 new tasks
    11/18/2012 7:27:35 PM | Einstein@Home | No work sent
    11/18/2012 7:27:35 PM | Einstein@Home | see scheduler log messages on http://einstein.phys.uwm.edu//host_sched_logs/6110/6110411
    11/18/2012 9:02:30 PM | Einstein@Home | Sending scheduler request: To fetch work.
    11/18/2012 9:02:30 PM | Einstein@Home | Requesting new tasks for NVIDIA
    11/18/2012 9:02:32 PM | Einstein@Home | Scheduler request completed: got 0 new tasks
    11/18/2012 9:02:32 PM | Einstein@Home | No work sent
    11/18/2012 9:02:32 PM | Einstein@Home | see scheduler log messages on http://einstein.phys.uwm.edu//host_sched_logs/6110/6110411
    11/18/2012 10:00:19 PM | | Windows is suspending operations
    11/18/2012 10:00:20 PM | | Suspending computation - requested by operating system
    11/18/2012 10:00:20 PM | | Suspending network activity - requested by operating system
    11/18/2012 10:00:21 PM | | Resuming after OS suspension
    11/18/2012 10:21:32 PM | | Windows is resuming operations
    11/18/2012 10:21:36 PM | | Resuming computation
    11/18/2012 10:21:36 PM | | Resuming network activity

    Jeroen
    Jeroen
    Joined: 25 Nov 05
    Posts: 379
    Credit: 740030628
    RAC: 0

    No work for GPU?

    19 Nov 2012 5:22:00 UTC
    Message 113058
    (moderation:
      )

      Hello,

      Login to your Einstein@Home account via this website.

      Under Computer Preferences, set the following:

      Suspend work while computer is in use? no
      Suspend GPU work while computer is in use? no

      Under Einstein@Home preferences, set the following:

      Use NVIDIA GPU: yes
      Run only the selected applications: Make sure there is a check for Binary Radio Pulsar Search (Arecibo)

      Go to your BOINC Manager and update the Einstein project and check to see if your system can get NVIDIA tasks.

      Gundolf Jahn
      Gundolf Jahn
      Joined: 1 Mar 05
      Posts: 1079
      Credit: 341280
      RAC: 0

      You are using the wrong

      19 Nov 2012 8:28:11 UTC
      Message 113059
      (moderation:
        )

        You are using the wrong graphic driver version (295.55), as can be seen in your last server contact log:

        Quote:
        2012-11-19 04:20:58.9831 [PID=10447] [version] Checking plan class 'BRP4cuda32'
        2012-11-19 04:20:58.9832 [PID=10447] [version] parsed project prefs setting 'gpu_util_brp': 0.000000
        2012-11-19 04:20:58.9832 [PID=10447] [version] driver version required max: -29053, supplied: 29555
        2012-11-19 04:20:58.9832 [PID=10447] [version] Checking plan class 'BRP4cuda32nv301'
        2012-11-19 04:20:58.9832 [PID=10447] [version] parsed project prefs setting 'gpu_util_brp': 0.000000
        2012-11-19 04:20:58.9832 [PID=10447] [version] driver version required min: -30100, supplied: 29555


        Gruß,
        Gundolf

        Computer sind nicht alles im Leben. (Kleiner Scherz)

        Richard
        Richard
        Joined: 29 Apr 10
        Posts: 39
        Credit: 39269698
        RAC: 0

        I think we are getting

        20 Nov 2012 12:40:00 UTC
        Message 113060 in response to message 113059
        (moderation:
          )

          I think we are getting closer. I updated the display driver, and it does find wok now, in fact it downloads 6 tasks every time. The only problem is as soon as it gets downloaded the tasks end with a computation error.

          mikebw
          mikebw
          Joined: 21 Nov 10
          Posts: 2
          Credit: 67831931
          RAC: 0

          I get the same computation

          20 Nov 2012 13:24:36 UTC
          Message 113061 in response to message 113060
          (moderation:
            )

            I get the same computation error, and I have to reset the project for it to get more work units after the failure. I have the latest CUDA Driver Version: 4.2.10, and I am running BOINC Manager 7.0.31 (x86) on my MacPro 10.6.8. nVidia GT120 (and ATI Radeon 5870).

            I had things setup to run rosetta@Home and Poem@Home on CPU and Einstein@Home on GPU only. This was working great for months, up until about Oct 30th where it seems to have stopped doing tasks for Einstein.

            Jord
            Joined: 26 Jan 05
            Posts: 2952
            Credit: 5893653
            RAC: 51

            RE: I think we are getting

            20 Nov 2012 15:34:17 UTC
            Message 113062 in response to message 113060
            (moderation:
              )

              Quote:
              I think we are getting closer. I updated the display driver, and it does find wok now, in fact it downloads 6 tasks every time. The only problem is as soon as it gets downloaded the tasks end with a computation error.


              Your error is "exit code -1073741515 (0xc0000135)", there's a thread about it here.

              However, seeing that you run one of the newer i7 CPUs, I suspect you installed drivers for it and that with those drivers Intel installed OpenCL capability for the built-in GPU on the Ivy Bridge processor (the HD Graphics 3000). Whether you use it or not...

              By installing the OpenCL components for the Intel GPU, parts of the AMD OpenCL code got overwritten. Not enough for the OpenCL detection to no longer work, but enough for it to work in a wonky way on your GPU, since the code that it now uses is optimized to be used on an Intel GPU, not an AMD GPU.

              It can easily be solved by uninstalling the Intel OpenCL drivers. It should show up in Uninstall Programs as something like Intel OpenCL Support. Uninstalling this will restore the original AMD OpenCL components.

              Jord
              Joined: 26 Jan 05
              Posts: 2952
              Credit: 5893653
              RAC: 51

              RE: I get the same

              20 Nov 2012 15:40:56 UTC
              Message 113063 in response to message 113061
              (moderation:
                )

                Quote:
                I get the same computation error


                No, you don't. You do get a computation error, but it's not the same as Richard did. Yours is a "process exited with code 234 (0xea, -22)" error. You can see these by clicking on the task name.

                stderr further says:

                Quote:
                [01:42:23][71056][INFO ] Application startup - thank you for supporting Einstein@Home!
                [01:42:23][71056][INFO ] Starting data processing...
                [01:42:34][71056][ERROR] Failed to enable CUDA thread yielding for device #0 (error: 702)! Sorry, will try to occupy one CPU core...
                [01:42:42][71056][ERROR] Couldn't acquire CUDA context of device #0 (error: 702)!
                [01:42:42][71056][ERROR] Demodulation failed (error: 1002)!
                01:42:42 (71056): called boinc_finish


                Seems to be something wrong with your Nvidia card.

                Have you tried a reboot of the system?

                mikebw
                mikebw
                Joined: 21 Nov 10
                Posts: 2
                Credit: 67831931
                RAC: 0

                Thanks Jord, restarting did

                21 Nov 2012 13:14:28 UTC
                Message 113064 in response to message 113063
                (moderation:
                  )

                  Thanks Jord, restarting did the trick.

                  Richard
                  Richard
                  Joined: 29 Apr 10
                  Posts: 39
                  Credit: 39269698
                  RAC: 0

                  I found a driver for "OpenCL

                  21 Nov 2012 14:53:11 UTC
                  Message 113065 in response to message 113062
                  (moderation:
                    )

                    I found a driver for "OpenCL 1.1 forIntel Core Processor Family" and uninstalled it. There is one task which is running for the Nvidia. I think that fixed it. Thanks for the help.

                    vjacob
                    vjacob
                    Joined: 11 Feb 05
                    Posts: 6
                    Credit: 0
                    RAC: 0

                    RE: I get the same

                    2 Dec 2012 9:36:48 UTC
                    Message 113066 in response to message 113061
                    (moderation:
                      )

                      Quote:

                      I get the same computation error, and I have to reset the project for it to get more work units after the failure. I have the latest CUDA Driver Version: 4.2.10, and I am running BOINC Manager 7.0.31 (x86) on my MacPro 10.6.8. nVidia GT120 (and ATI Radeon 5870).

                      I had things setup to run rosetta@Home and Poem@Home on CPU and Einstein@Home on GPU only. This was working great for months, up until about Oct 30th where it seems to have stopped doing tasks for Einstein.

                      I am not getting that error, but I am also not sure if I am getting tasks for the GPU(s).
                      - How can one tell from the tasks? Unlike PrimeGrid, there is no parenthesis indicating the type of processing approach, i.e. "(cuda32)", and the "(SSE2)" or "(BRP4SSE)" does not tell me..?

                      FYI: BOINC client version 7.0.31 for x86_64-apple-darwin on my MacBook Pro, 8Gb physical memory, a GeForce 9600M GT + 9400M, CUDA 5.0.37 (just downloaded a few days ago). My GPU driver is listed as 1.6.36.10 (256.00.35f11).

                      NVIDIA/OpenCL/CUDA lines from the log:
                      Sun Dec 2 06:40:31 2012 | | NVIDIA GPU 0: GeForce 9600M GT (driver version 5.0.37, CUDA version 5.0, compute capability 1.1, 256MB, 53MB available, 120 GFLOPS peak)
                      Sun Dec 2 06:40:31 2012 | | NVIDIA GPU 1: GeForce 9400M (driver version 5.0.37, CUDA version 5.0, compute capability 1.1, 254MB, 231MB available, 53 GFLOPS peak)
                      Sun Dec 2 06:40:31 2012 | | OpenCL: NVIDIA GPU 0: GeForce 9600M GT (driver version CLH 1.0, device version OpenCL 1.0, 256MB, 53MB available)
                      Sun Dec 2 06:40:31 2012 | | OpenCL: NVIDIA GPU 1: GeForce 9400M (driver version CLH 1.0, device version OpenCL 1.0, 256MB, 231MB available)

                      Task events from the BOINC log:
                      Sun Dec 2 09:27:51 2012 | | Contacting account manager at http://www.gridrepublic.org/
                      Sun Dec 2 09:27:54 2012 | | Account manager contact succeeded
                      Sun Dec 2 09:27:56 2012 | Einstein@Home | Sending scheduler request: Requested by account manager.
                      Sun Dec 2 09:27:56 2012 | Einstein@Home | Requesting new tasks for CPU and NVIDIA
                      Sun Dec 2 09:28:03 2012 | Einstein@Home | Scheduler request completed: got 2 new tasks
                      Sun Dec 2 09:29:04 2012 | Einstein@Home | Sending scheduler request: To fetch work.
                      Sun Dec 2 09:29:04 2012 | Einstein@Home | Requesting new tasks for CPU and NVIDIA
                      Sun Dec 2 09:29:09 2012 | Einstein@Home | Scheduler request completed: got 1 new tasks
                      Sun Dec 2 09:29:11 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_80.bin4
                      Sun Dec 2 09:29:11 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_81.bin4
                      Sun Dec 2 09:29:24 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_81.bin4
                      Sun Dec 2 09:29:24 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_82.bin4
                      Sun Dec 2 09:29:25 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_80.bin4
                      Sun Dec 2 09:29:25 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_83.bin4
                      Sun Dec 2 09:29:29 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_82.bin4
                      Sun Dec 2 09:29:29 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_84.bin4
                      Sun Dec 2 09:29:31 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_83.bin4
                      Sun Dec 2 09:29:31 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_85.bin4
                      Sun Dec 2 09:29:39 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_84.bin4
                      Sun Dec 2 09:29:39 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_86.bin4
                      Sun Dec 2 09:29:46 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_86.bin4
                      Sun Dec 2 09:29:46 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_87.bin4
                      Sun Dec 2 09:29:54 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_87.bin4
                      Sun Dec 2 09:29:54 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000.zap
                      Sun Dec 2 09:29:55 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000.zap
                      Sun Dec 2 09:29:58 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b6s0g0.00000_85.bin4
                      Sun Dec 2 09:30:09 2012 | Einstein@Home | Sending scheduler request: To fetch work.
                      Sun Dec 2 09:30:09 2012 | Einstein@Home | Requesting new tasks for CPU and NVIDIA
                      Sun Dec 2 09:30:13 2012 | Einstein@Home | Scheduler request completed: got 1 new tasks
                      Sun Dec 2 09:30:16 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1824.bin4
                      Sun Dec 2 09:30:16 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1825.bin4
                      Sun Dec 2 09:30:21 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1825.bin4
                      Sun Dec 2 09:30:21 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1826.bin4
                      Sun Dec 2 09:30:24 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1824.bin4
                      Sun Dec 2 09:30:24 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1827.bin4
                      Sun Dec 2 09:30:26 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1826.bin4
                      Sun Dec 2 09:30:26 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1828.bin4
                      Sun Dec 2 09:30:28 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1827.bin4
                      Sun Dec 2 09:30:28 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1829.bin4
                      Sun Dec 2 09:30:34 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1828.bin4
                      Sun Dec 2 09:30:34 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1829.bin4
                      Sun Dec 2 09:30:34 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1830.bin4
                      Sun Dec 2 09:30:34 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1831.bin4
                      Sun Dec 2 09:30:40 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1830.bin4
                      Sun Dec 2 09:30:40 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000.zap
                      Sun Dec 2 09:30:41 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000_1831.bin4
                      Sun Dec 2 09:30:41 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b0s0g0.00000.zap
                      Sun Dec 2 09:31:14 2012 | Einstein@Home | Sending scheduler request: To fetch work.
                      Sun Dec 2 09:31:14 2012 | Einstein@Home | Requesting new tasks for CPU and NVIDIA
                      Sun Dec 2 09:31:19 2012 | Einstein@Home | Scheduler request completed: got 1 new tasks
                      Sun Dec 2 09:31:21 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2784.bin4
                      Sun Dec 2 09:31:21 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2785.bin4
                      Sun Dec 2 09:31:28 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2784.bin4
                      Sun Dec 2 09:31:28 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2786.bin4
                      Sun Dec 2 09:31:33 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2786.bin4
                      Sun Dec 2 09:31:33 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2787.bin4
                      Sun Dec 2 09:31:35 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2785.bin4
                      Sun Dec 2 09:31:35 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2788.bin4
                      Sun Dec 2 09:31:40 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2787.bin4
                      Sun Dec 2 09:31:40 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2789.bin4
                      Sun Dec 2 09:31:45 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2788.bin4
                      Sun Dec 2 09:31:45 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2790.bin4
                      Sun Dec 2 09:31:49 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2790.bin4
                      Sun Dec 2 09:31:49 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2791.bin4
                      Sun Dec 2 09:31:50 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2789.bin4
                      Sun Dec 2 09:31:50 2012 | Einstein@Home | Started download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000.zap
                      Sun Dec 2 09:31:51 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000.zap
                      Sun Dec 2 09:31:52 2012 | Einstein@Home | Finished download of p2030.20111230.G176.57-03.05.S.b1s0g0.00000_2791.bin4

                      Best
                      J

                      Claggy
                      Claggy
                      Joined: 29 Dec 06
                      Posts: 560
                      Credit: 2699403
                      RAC: 0

                      RE: RE: I get the same

                      2 Dec 2012 10:31:50 UTC
                      Message 113067 in response to message 113066
                      (moderation:
                        )

                        Quote:
                        Quote:

                        I get the same computation error, and I have to reset the project for it to get more work units after the failure. I have the latest CUDA Driver Version: 4.2.10, and I am running BOINC Manager 7.0.31 (x86) on my MacPro 10.6.8. nVidia GT120 (and ATI Radeon 5870).

                        I had things setup to run rosetta@Home and Poem@Home on CPU and Einstein@Home on GPU only. This was working great for months, up until about Oct 30th where it seems to have stopped doing tasks for Einstein.

                        I am not getting that error, but I am also not sure if I am getting tasks for the GPU(s).
                        - How can one tell from the tasks? Unlike PrimeGrid, there is no parenthesis indicating the type of processing approach, i.e. "(cuda32)", and the "(SSE2)" or "(BRP4SSE)" does not tell me..?

                        FYI: BOINC client version 7.0.31 for x86_64-apple-darwin on my MacBook Pro, 8Gb physical memory, a GeForce 9600M GT + 9400M, CUDA 5.0.37 (just downloaded a few days ago). My GPU driver is listed as 1.6.36.10 (256.00.35f11).

                        NVIDIA/OpenCL/CUDA lines from the log:
                        Sun Dec 2 06:40:31 2012 | | NVIDIA GPU 0: GeForce 9600M GT (driver version 5.0.37, CUDA version 5.0, compute capability 1.1, 256MB, 53MB available, 120 GFLOPS peak)
                        Sun Dec 2 06:40:31 2012 | | NVIDIA GPU 1: GeForce 9400M (driver version 5.0.37, CUDA version 5.0, compute capability 1.1, 254MB, 231MB available, 53 GFLOPS peak)
                        Sun Dec 2 06:40:31 2012 | | OpenCL: NVIDIA GPU 0: GeForce 9600M GT (driver version CLH 1.0, device version OpenCL 1.0, 256MB, 53MB available)
                        Sun Dec 2 06:40:31 2012 | | OpenCL: NVIDIA GPU 1: GeForce 9400M (driver version CLH 1.0, device version OpenCL 1.0, 256MB, 231MB available)


                        Can you post a link to this host, with your hosts hidden no-one can help you much,

                        Claggy

                        Comment viewing options

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

                        Forums › Problems and Bug Reports