Announcement

Collapse
No announcement yet.

AMD Confirms Linux Performance Marginality Problem Affecting Some, Doesn't Affect Epyc / TR

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Originally posted by InsideJob View Post
    I have to admit I didn't read half the posts in this thread -- you're all embarrassing yourselves with public displays of ignorance -- but I'm not paying for "performance marginality" when they sold me XFR.

    I smell another class-action lawsuit in the making. That'll learn 'em real good!
    *lol* You're really angry about this, aren't you?

    Comment


    • Is there a bug in the ignore list function? I just added duby229 and sdack to mine, but still suffer having to filter their noise.

      Comment


      • Originally posted by Beherit View Post
        Is there a bug in the ignore list function? I just added duby229 and sdack to mine, but still suffer having to filter their noise.
        I'll stop. Please, just don't threaten Phoronix with a class action lawsuit for the ignore function not working as expected.
        Last edited by sdack; 08 August 2017, 12:24 PM.

        Comment


        • Originally posted by Beherit View Post
          Is there a bug in the ignore list function? ...
          Yes.
          Unfortunately for a very long time. :-(

          Comment


          • Wendel is saying the same issue is present on the newest Intel systems, but it's just in conftest. So I guess we'll let them sort it out.

            Comment


            • Originally posted by pjssilva View Post

              Actually you can do this kind of stuff by suing a configuration file in /etc/sensors ou /etc/sensors.d but I do not remember the details. In this case the formula is no easy that I do it in my head. Note that the -20 C is only necessary for the X processors, like 1700X and 1800X. For a regular processor, like 1700, the temperature will be already correct. I forgot to mention that. Sorry.
              I just did it for my setup. Here is an example how to do it (except I think they set second value incorrectly as negative, it should be reversed): https://wiki.archlinux.org/index.php...rature_offsets

              First check your chip and sensor you want to adjust:

              sensors -u
              Code:
              nct6779-isa-0290
              ...
              SMBUSMASTER 0:
                temp7_input: 48.000
              Then create something like this: /etc/sensors.d/asrock-x370-taichi.conf

              Code:
              chip "nct6779-isa-0290"
                compute temp7 @-20,@+20
              Now SMBUSMASTER 0 reports correct CPU temperature:

              sensors
              Code:
              ...
              SMBUSMASTER 0:          +28.0°C
              You can also see man sensors.conf

              For me the problem is that KDE thermal monitor plasmoid is ignoring these offsets.
              Last edited by shmerl; 08 August 2017, 12:55 PM.

              Comment


              • I would think twice about believing a company that has taken from May until now to even own up
                to the fact there was something wrong with their Ryzen chip, I read 39 pages on the AMD forum
                of complaints, and endless examples of how to re create the problem.

                I will wait until Phoronix tests the Threadripper, and the community subject it to the same tests
                that showed the Ryzen for what it is, until then, I for One do not believe AMD, as I think it is no coincidence
                that they only owned up right before the release of their Threadripper, and in doing so got sell the amazing
                piece of news to us all that Threadripper is not sick like the Ryzen.

                Comment


                • Originally posted by sdack View Post
                  *lol* I'm no hypocrite, but you're a fear-mongerer. First you want AMD to confirm your worst fears, and when it's not happening, but they rather tell you it doesn't affect Windows and it's only a rare case found only under Linux, which we knew all along, do you choose not to believe AMD, but want to hold on to your fears. Very smooth of you ... *lol*

                  If I'm the one sitting in a cave then it sure isn't as dark as yours.
                  You're really aiming for debianxfce's crown, aren't you? Not an easy task, but I'm impressed with your performance so far.

                  Comment


                  • Originally posted by Windingroads View Post
                    I would think twice about believing a company that has taken from May until now to even own up
                    to the fact there was something wrong with their Ryzen chip, I read 39 pages on the AMD forum
                    of complaints, and endless examples of how to re create the problem.

                    I will wait until Phoronix tests the Threadripper, and the community subject it to the same tests
                    that showed the Ryzen for what it is, until then, I for One do not believe AMD, as I think it is no coincidence
                    that they only owned up right before the release of their Threadripper, and in doing so got sell the amazing
                    piece of news to us all that Threadripper is not sick like the Ryzen.
                    You do realize that intel has done the same thing multiple times in the past? ISTR that they sat on the Skylake HT crash bug for over a year.

                    This kind of behaviour is par for the course in this particular game I'm afraid.

                    Comment


                    • Originally posted by pjssilva View Post
                      The CPU temperature is in SMBUSMASTER (with a 20 C offset).
                      Isn't this offset is only needed for internal CPU thermal sensor? I would expect that Super I/O chip is already showing the correct value.

                      Comment

                      Working...
                      X