Announcement

Collapse
No announcement yet.

Trade in Quadro K4000 for 2X GTX970

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

  • #16
    Originally posted by vlado View Post
    Well, the K6000 also has 12 GB of VRAM I don't know how to measure the viewport performance though - any pointers?

    Best regards,
    Vlado
    How about this scene?
    Saw it in Evermotion years ago.

    City01-V1d.zip
    http://depositfiles.com/files/qnwmfhn0m
    This is a signature.

    Comment


    • #17
      Hmpf, which download button do I need to press on that page?

      Best regards,
      Vlado
      I only act like I know everything, Rogers.

      Comment


      • #18
        You need to get adblock - any advertiser using those sites now just makes ads that say 'download now' so you dont know what's real or not.

        It's the big red 'regular download' button underneath the speedometer (the one that looks most likely to be a fake link, yes...)
        Then there's a catchpa, another button that pops up, a 5 second countdown after you click that and you're good to go.

        Comment


        • #19
          https://www.dropbox.com/s/v7gxda36rr...1-V1d.rar?dl=0

          uploaded to dropbox.
          www.archcg.my

          Comment


          • #20
            Should be an interesting test I'll try that on my Quadro K4000 and report back.
            A.

            ---------------------
            www.digitaltwins.be

            Comment


            • #21
              Nitrous with DX11 gives 50+ fps on a GTX 980.

              Best regards,
              Vlado
              I only act like I know everything, Rogers.

              Comment


              • #22
                Tested this on my work on a regular Quadro 5000 (MAX 2014, nitrous DX11), in the beginning (first 250 frames) I have an FPS of +- 20, then it goes up to around 35-50 till the end. Will test on Quadro K4000 tonight.
                A.

                ---------------------
                www.digitaltwins.be

                Comment


                • #23
                  As far as I remember, the folks at Evermotion forum measured time, not FPS.
                  Should have mentioned that earlier.

                  About FPS, I don't understand - it's the nitrous improved in 2015 or the new GPU is that powerful.
                  This is a signature.

                  Comment


                  • #24
                    Originally posted by 3dcrgr View Post
                    About FPS, I don't understand - it's the nitrous improved in 2015 or the new GPU is that powerful.
                    How do I measure the time?

                    I will need to replace the graphics card to see if the improvements are due to Nitrous or not, but it won't be soon.

                    Best regards,
                    Vlado
                    I only act like I know everything, Rogers.

                    Comment


                    • #25
                      Originally posted by vlado View Post
                      How do I measure the time?

                      I will need to replace the graphics card to see if the improvements are due to Nitrous or not, but it won't be soon.

                      Best regards,
                      Vlado
                      I never asked how they did it, but I guess most of them just hit play in max and a stop watch at the same time.
                      As soon as the sequence returned to 0, you stop measuring.
                      Test were done in wireframe, shaded and hidden line.

                      At that time They claimed pan/rotate/zoom the scene is not very accurate and based on personal estimations.
                      While with a stop watch and a camera defined movement you can get a more accurate result.

                      I will post later result of MAX 2014 with GTX 770 and GTX 780.
                      This is a signature.

                      Comment


                      • #26
                        Originally posted by 3dcrgr View Post
                        I never asked how they did it, but I guess most of them just hit play in max and a stop watch at the same time.
                        But the playback goes through that in a few seconds, how can it be timed precisely?

                        Actually I have 3ds Max 2014 too, I can compare the performance between 2014 and 2015.

                        Best regards,
                        Vlado
                        I only act like I know everything, Rogers.

                        Comment


                        • #27
                          Originally posted by vlado View Post
                          But the playback goes through that in a few seconds, how can it be timed precisely?

                          Actually I have 3ds Max 2014 too, I can compare the performance between 2014 and 2015.

                          Best regards,
                          Vlado
                          I timed it with my phone and clicked while i released my phone-timer at the same time. It might be off for like a few hundredths of a second.

                          This is the original thread: http://www.evermotion.org/vbulletin/...ty01-max-File). It's a little outdated but a good benchmark I guess nonetheless. Although if you only test it in wireframe without any textures and stuff, I don't know if it's such a good one.

                          Well the results in wireframe mode for my Quadro K4000 is not that much different as for the regular Quadro 5000, FPS is around 25 -30 fps in the beginning, when most geometry is in view.

                          The complete animation is done in around 17.66 seconds.
                          A.

                          ---------------------
                          www.digitaltwins.be

                          Comment


                          • #28
                            Originally posted by Vizioen View Post
                            I timed it with my phone and clicked while i released my phone-timer at the same time. It might be off for like a few hundredths of a second.

                            This is the original thread: http://www.evermotion.org/vbulletin/...ty01-max-File). It's a little outdated but a good benchmark I guess nonetheless. Although if you only test it in wireframe without any textures and stuff, I don't know if it's such a good one.

                            Well the results in wireframe mode for my Quadro K4000 is not that much different as for the regular Quadro 5000, FPS is around 25 -30 fps in the beginning, when most geometry is in view.

                            The complete animation is done in around 17.66 seconds.
                            It is a short time.
                            I remember it took a minute and a half on older hardware, and older versions of Max.
                            This is a signature.

                            Comment

                            Working...
                            X