Discussion Thread for the Continuous GW Search known as O2MD1 (now O2MDF - GPUs only)

Forums › Cruncher's Corner

Richie
Richie
Joined: 7 Mar 14
Posts: 656
Credit: 1702989778
RAC: 0

Bernd Machenschalk wrote:The

8 Feb 2021 19:44:32 UTC
Message 183301 in response to message 183292
(moderation:
    )

    Bernd Machenschalk wrote:
    The CPU part of the run ("O2MD1S3") is also running now.

    Great! There's still a glitch on the flow though... Downloads are failing. Success rate 0/10. Permanent HTTP errors.

    https://einsteinathome.org/task/1069674861

    Tom M
    Tom M
    Joined: 2 Feb 06
    Posts: 6461
    Credit: 9586354537
    RAC: 6877232

    Thank you for the update

    8 Feb 2021 19:49:07 UTC
    Message 183302 in response to message 183292
    (moderation:
      )

      Thank you for the update Bernd.

      Tom M

      A Proud member of the O.F.A.  (Old Farts Association).  Be well, do good work, and keep in touch.® (Garrison Keillor)  I want some more patience. RIGHT NOW!

      Bernd Machenschalk
      Bernd Machenschalk
      Moderator
      Administrator
      Joined: 15 Oct 04
      Posts: 4312
      Credit: 250603337
      RAC: 34728

      Richie wrote: Bernd

      8 Feb 2021 20:56:54 UTC
      Message 183304 in response to message 183301
      (moderation:
        )

        Richie wrote:

        Bernd Machenschalk wrote:
        The CPU part of the run ("O2MD1S3") is also running now.

        Great! There's still a glitch on the flow though... Downloads are failing. Success rate 0/10. Permanent HTTP errors.

        https://einsteinathome.org/task/1069674861

        Oh dear! Thanks for the note! Shouldn't take long to fix. Sending GW work is suspended until this is fixed.

        BM

        Tom M
        Tom M
        Joined: 2 Feb 06
        Posts: 6461
        Credit: 9586354537
        RAC: 6877232

        Does anyone have a clue about

        11 Feb 2021 14:05:24 UTC
        Message 183387
        (moderation:
          )

          Does anyone have a clue about the highest performing GW GPU tasks system?

          I am in the process of working up a GW only system and am wondering what kind of goal I should be shooting for.

          Tom M

          A Proud member of the O.F.A.  (Old Farts Association).  Be well, do good work, and keep in touch.® (Garrison Keillor)  I want some more patience. RIGHT NOW!

          Tom M
          Tom M
          Joined: 2 Feb 06
          Posts: 6461
          Credit: 9586354537
          RAC: 6877232

          I just got done bumping this

          14 Feb 2021 4:20:47 UTC
          Message 183447
          (moderation:
            )

            I just got done bumping this system up to 6 gpus with 2 threads per GPU.  I wanted to keep as many cpus available for additional expansion of GPU cards as I could.  So I set the limit to 0.5 CPU per GPU thread.

            The preliminary result was each thread was taking twice as long as it had previously.  This means there was no gain in running two threads per GPU. 

            I have since reset the CPU to the original default (0.9) per GPU thread.  I thought I saw a speed up in massive speed up in processing. 

            ==edit===

            It maybe that I am now processing at the same speed as a single thread per GPU.

            ====edit=deleted===

            There probably is a hard limit on the number of GPU threads a single system can push on GW tasks. No more that 1 gpu task per thread.  Which means you likely can't run 38 GPU threads at "full speed" on a 18 slot MB like the B360-F Pro because it has an 8c/16t limit.

            Tom M

            A Proud member of the O.F.A.  (Old Farts Association).  Be well, do good work, and keep in touch.® (Garrison Keillor)  I want some more patience. RIGHT NOW!

            Cherokee150
            Cherokee150
            Joined: 13 May 11
            Posts: 24
            Credit: 885460023
            RAC: 330585

            Does anyone know why there

            1 May 2021 14:12:22 UTC
            Message 185392
            (moderation:
              )

              Does anyone know why there are no new work units available for O2MDF?

              Ian&Steve C.
              Ian&Steve C.
              Joined: 19 Jan 20
              Posts: 3960
              Credit: 47071362642
              RAC: 65410956

              as far as I know they are

              1 May 2021 14:32:34 UTC
              Message 185393
              (moderation:
                )

                as far as I know they are working on the transition to O3ASE. O2MDF will go away.

                _________________________________________________________________________

                Cherokee150
                Cherokee150
                Joined: 13 May 11
                Posts: 24
                Credit: 885460023
                RAC: 330585

                Thank you. I look forward

                1 May 2021 15:08:16 UTC
                Message 185394 in response to message 185393
                (moderation:
                  )

                  Thank you.

                  I look forward to O3ASE.

                  I sincerely hope they program O3ASE to run on GPUs with only 2 GB of RAM.  I would like to be able to run them not only on my computer with an NVIDIA 1070, but also on my computer with an NVIDIA GeForce GTX 960.  There are still a lot of older GPUs out here with 2 GB of RAM that can process a lot of work units per day for Einstein. :-)

                  petri33
                  petri33
                  Joined: 4 Mar 20
                  Posts: 124
                  Credit: 4062195819
                  RAC: 6847518

                  Hi Bernd, I looked at my

                  1 May 2021 16:03:18 UTC
                  Message 185395 in response to message 183304
                  (moderation:
                    )

                    Hi Bernd,

                    I looked at my .nv/ComputeCache and found some new OpenCL code in the source that has Xlal_BSGL stuff and SemiCoherent code

                    <code>

                    ... this is just like it was before ..

                    #ifndef PULSAR_MAX_DETECTORS

                    #define PULSAR_MAX_DETECTORS 2

                    #endif

                    ... and a couple of lines below is some new code, that is overidden by the previous definition and has no effect

                    #ifndef PULSAR_MAX_DETECTORS

                    #define PULSAR_MAX_DETECTORS 10

                    #endif

                    </code>

                    a) Is this the way it should be?

                    b) Another question: Is there a known hard upper limit in CopyBSGLSetup() for the UINT4 numDetectors argument ?

                    petri33
                    petri33
                    Joined: 4 Mar 20
                    Posts: 124
                    Credit: 4062195819
                    RAC: 6847518

                      2021-05-03

                    3 May 2021 8:43:15 UTC
                    Message 185432
                    (moderation:
                      )

                       

                      2021-05-03 10:54:25.9746 (205377) [normal]: Recalc FstatMethod used: 'DemodSSE'

                      O3ASE tasks spend about three minutes recalculating statistics on CPU after main analysyis on GPU has finished:

                      2021-05-03 10:57:57.5918 (205377) [normal]: Finished main analysis.
2021-05-03 10:57:57.5921 (205377) [normal]: Recalculating statistics for the final toplist...
2021-05-03 11:00:48.1217 (205377) [normal]: Finished recalculating toplist statistics.

                      The O2MD did the same much faster (50 seconds)


                      2021-03-18 22:42:18.3670 (214468) [normal]: Finished main analysis.
2021-03-18 22:42:18.3672 (214468) [normal]: Recalculating statistics for the final toplist...
2021-03-18 22:43:11.3893 (214468) [normal]: Finished recalculating toplist statistics.

                      Please do not take this too seriously. I'm just wondering:

                      Is this recalculation just to check if GPU has done something right? It says re calculating. Is it necessary at all?

                      The GPU sits three minutes idle. Looks bad: 20 s initial verification + 200 s work + 180 seconds idle.

                      I know, I could run two at a time, but NVIDIA GPUS slow down a lot when doing that. Oh how I miss the Seti mutex implementation that let the other task do GPU-stuff when the other had finished its GPU part. Pre- and post-steps on CPU overlapped nicely with the other task that was doing GPU.

                       

                      Petri33

                      Comment viewing options

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

                      Forums › Cruncher's Corner