9 Replies Latest reply on Apr 1, 2013 2:55 AM by Harm Millaard

    Does PPBM6 work properly with Adobe's free trial CS6 download?

    JEShort01 Level 4
      1. Does PPBM6 work properly with Adobe's free trial download of Premiere Pro CS6?
      2. Does the CS6 download include the 6.0 level release, or is it the latest version (or at least download patchable to the latest version)?
      3. If anyone has actually done exactly this, please share your experience?

       

       

       

      Thanks,

       

      Jim

        • 1. Re: Does PPBM6 work properly with Adobe's free trial CS6 download?
          Bill Gehrke Adobe Community Professional & MVP

          Jim,

           

          1. I cannot see any reason why our new PPBM6 would not work.
          2. Do not know that answer
          3. We do have a few results posted if you have logged into our new site.
          • 2. Re: Does PPBM6 work properly with Adobe's free trial CS6 download?
            RjL190365 Level 4

            Bill Gehrke wrote:

             

            We do have a few results posted if you have logged into our new site.

             

            I just saw it. Out of the nine results so far, a pattern is beginning to emerge outside of the obvious disk performance scores:

             

            1. Systems with hexa-core CPUs consistently had better (faster) H.264 Blu-ray encoding scores than systems with quad-core CPUs.
            2. Among systems based on LGA 1155 CPUs, those with i7 CPUs outperformed those with i5 CPUs in that same H.264 Blu-ray encoding test.
            3. Systems with the faster-performing CPUs generally had lower MPE performance gain factors than those with lesser-performing CPUs.
            4. CPU clock speed has a moderate impact on the H.264 scores while memory speed has a much smaller impact on those same scores.
            • 3. Re: Does PPBM6 work properly with Adobe's free trial CS6 download?
              Bill Gehrke Adobe Community Professional & MVP

              RjL190365 wrote:

              I just saw it. Out of the nine results so far, a pattern is beginning to emerge outside of the obvious disk performance scores:

               

              1. Systems with hexa-core CPUs consistently had better (faster) H.264 Blu-ray encoding scores than systems with quad-core CPUs.
              2. Among systems based on LGA 1155 CPUs, those with i7 CPUs outperformed those with i5 CPUs in that same H.264 Blu-ray encoding test.
              3. Systems with the faster-performing CPUs generally had lower MPE performance gain scores than those with lesser-performing CPUs.
              4. CPU clock speed has a moderate impact on the H.264 scores while memory speed has a much smaller impact on those same scores.

              In general your conclusions are pretty much correct, but without that data being posted yet and only 9 samples I would have a hard time proving your statements 

              • 4. Re: Does PPBM6 work properly with Adobe's free trial CS6 download?
                RjL190365 Level 4

                That's why I meant that to be preliminary.

                 

                And GPU generation has relatively little impact on the MPEG-2 DVD scores: The GTX 560 Ti 448 proved competitive with newer, higher-end GPUs in that test in terms of encoding time, but I think the relatively slow performance of the plain GTX 560 in my "Randall's Mud Rover" system might have been due in part to the system having only 16GB of RAM (as opposed to the 32GB or more RAM in most of the other systems posted thus far).

                • 5. Re: Does PPBM6 work properly with Adobe's free trial CS6 download?
                  Harm Millaard Level 7

                  Randall,

                   

                  Just a snippet from my mail to Bill about testing results up to now:

                   

                   

                  MPEG2 results appear to be dependent not on number of cores, but on amount of memory, next to the video card and maybe memory controller. How else would your SB be faster than your 980 with the same video card and 4 cores versus 6? I think 32 GB versus 24 GB is the differentiating factor here, as well as the memory controller on the SB, or am I trying to find explanations that are not there?

                   

                  • Disk I/O is purely a matter of sustained transfer rates. Number of cores is irrelevant here. I don't have sufficient data yet to determine whether amount of memory will be influential, but I assume there is a lower limit when the results will further decrease. We are talking about a 37,092 MB file being written to disk and the sustained transfer rate can be negatively influenced by a limited amound of memory.
                  • MPEG2-DVD as shown above.
                  • H.264 is mainly a factor of number of cores, size of the L3 cache, amount and speed of the memory and clock speed.
                  • MPE Gain is a resultant calculated number. It is hard to interpret, I admit, but in general you will see that the better a system is performing, the lower the Gain figure. Very fast systems can handle software MPEG2-DVD encoding quite well, slower systems struggle with that task, but with hardware MPE even slower systems achieve around the same figure as faster systems and the consequence is that slower systems show a higher gain.

                   

                  Further notice that the new Total Time figure is only the sum of Disk I/O + MPEG2-DVD + H.264-BR. In the PPBM5 test we included the rendering test as well, but since we no longer use that, it is no longer a part of the Total score. Also, the weighing of test results to calculate the RPI has been adjusted for this new test.

                   

                  Hope this gives some informative background info.

                  • 6. Re: Does PPBM6 work properly with Adobe's free trial CS6 download?
                    RjL190365 Level 4

                    Thanks, Harm. Now I know why an i5 performs slower than an equally clocked LGA 1155 i7, not only because of the lack of Hyperthreading, but also because of the smaller L3 cache (6MB versus 8MB).

                    • 7. Re: Does PPBM6 work properly with Adobe's free trial CS6 download?
                      Harm Millaard Level 7

                      Randall, still awaiting your reply to my mail to your regular address about PPBM6.

                      • 8. Re: Does PPBM6 work properly with Adobe's free trial CS6 download?
                        RjL190365 Level 4

                        Bill Gehrke wrote:

                         

                        In general your conclusions are pretty much correct, but without that data being posted yet and only 9 samples I would have a hard time proving your statements

                         

                        There are now 11 results; however, two of those 11 are the same system retested and another two are of the same system with an older-generation GPU (GTX 470) and after an upgrade to a GTX 660.

                         

                        And looking at the XML results, it turned out that I wasn't the only person who submitted a result from a PC with only 16GB of RAM. I think Paul's system might have even more background processes running than I do, which is why his results running an i7-3770K overclocked to 4.4GHz with memory running at DDR3-1600 speed actually performed slower in the H.264 test than my system running the same i7-3770K CPU overclocked to 100MHz lower (4.3GHz) and running its RAM at only DDR3-1333 speed. (Though that auxiliary system of mine performed slowest in the MPEG-2 DVD test due to its use of the slowest GPU of the bunch that was submitted thus far.)

                        • 9. Re: Does PPBM6 work properly with Adobe's free trial CS6 download?
                          Harm Millaard Level 7

                          With the step I made yesterday, inspecting the differences between systems is now a lot easier. Instead of looking at hard to read .xml files, several early adopters have submitted their Speccy snapshot files (instructions and submission form have been updated to reflect this new move) and hopefully the rest will follow suit shortly.

                           

                          The results page now looks like this:

                          Speccy results.png

                          After opening a snapshot file and moving the windows around a bit you can easily compare multiple systems in detail.

                          Speccy details.png

                          and with sufficient screen space even like this:

                          Speccy details 3.png

                          This should be a lot easier than those .xml files.