31 Replies Latest reply on Jul 13, 2011 1:16 AM by Crist OC/PC

    How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score

    JEShort01 Level 4

      I have run PPBM5 44 times for various combinations of CPU speed, hard disk speed, and RAM on the same CPU/motherboard and will report specific conclusions in an effort to help fellow users build cost effective PCs to run CS5. I have gained a lot from this forum, but have posted less than 20 times. I guess one could say this is my feeble attempt to pay it forward.

       

      I'd like to thank first and foremost Bill and Harm for providing and supporting the PPBM5 benchmark. This provides an excellent common reference for everyone that measures different aspects of Premiere Pro CS5 performance. Next I'd like to thank Eric Bowen with ADK whose guidance has helped me a lot; many companies that build PCs for a living do not participate at all on forums. Finally, thanks to the many other contributors to this forum - there are too many to name!

       

      So what have I learned and how can it be communicated here is a way that is useful?

       

      I'm going to report specific PPBM5 results that reflect various hardware changes to a common motherboard and CPU and will focus on those results that significantly changed performance. Harm preaches the truth that a fast system requires a full complement of sub-systems; this is indeed what I've found.

       

      When I comment on:

      CPU speed - RAM size and drive performance has already been maximized,

      RAM size - CPU and drive performance has already been maximized,

      Drives - CPU speed and RAM size has already been maximized.

       

      Base system used for testing:

      Gigabyte GA-P55-UD4P, Intel i7-860 CPU, MSI Hawk GTX 460 1GB video @ 850MHz, OS/boot array is 2x80GB Intel X25-M SSD in RAID 0 (Win7 64-bit OS, CS5 v5.0.3)

       

      1) CPU speed; important and surprisingly affects all 4 PPBM5 tests, including the GPU MPE test; I saw very linear improvements going from 2.8GHz (stock) up to 4.0GHz (OC) [PPBM5 composite score: 268 seconds stock, 195 seconds at OC]

       

      2) RAM size: increasing from 4GB to 8GB doubled the MPE GPU performance [11 sec. vs. 6 sec.] and improved the MPEG2-DVD performance [122 sec. vs. 91 sec.]; the other tests only improved marginally

       

      Increasing from 8GB to 16GB made significant additional improvements to the MPEG2-DVD time [91 sec. vs. 37 sec.]; all other tests hardly changed at all

       

      3) Drives: Slowest tested was project, cache, and cache DB all sharing a single 1TB WD 7200rpm drive [PPBM5 Disk I/O test = 138 sec.] Changing to two of the same drives in a RAID 0 configuration was a massive improvement [PPBM5 Disk I/O test = 82 sec.] Bleeding edge drive arrays running on Areca controllers produced the fastest result, but were only 7 seconds faster than the 2x1TB RAID 0 array [PPBM5 Disk I/O test = 75 sec.]; a 5x64GB Crucial SSD Raid 0 array performing similar to a 8x1TB WD 7200rpm RAID 5 array.

       

      SSD sidebar: Using a 2xSSD (Intel X25-M) RAID 0 array for the PPBM5 project was slower than the 2x1TB 7200rpm setup [93 sec. vs. 82 sec.], however I will be using a SSD array in my upcoming build for the OS and programs. They just make the system feel more eager and responsive, similar to the way a good handling car feels more responsive than a luxury car or a truck. Note too that all SSDs are NOT created equal. My small (64GB and 80GB) SSDs have lightning quick reads and slow writes (250MB/s reads, 70MB/s writes), whereas the 256GB EXPENSIVE models often write 4 times the speed that mine do. Good 1TB 7200 hard drives tend to read and write at around 130MB/s.

       

      Regards,

       

      Jim

        • 1. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
          Bill Gehrke Adobe Community Professional & MVP

          Jim, great summary. This is one very good example of one of the prime purposes of this benchmark, to optimize what ever you have.

           

          Congratulations you have the highest scoring LGA 1156 board based system on the test results.  On your submitted data in the PPBM5 Results page all your results fall into the Q3 catagory which indicates you have a very well balanced configuration.

           

          Now you can see why we have been recommending the even slightly faster X58 class processors, more than 12 GB of RAM, motherboards (LGA1366) and do not generally recommend SSD drives.

           

          Thank you.

           

          Bill

          • 2. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
            JEShort01 Level 4

            Bill,

             

            The main reason that I did a lot of testing on the this LGA 1156 board was so that I can build a new X58 system (it will become my primary home PC) and I want to do so WITHOUT having to do much messing around with it. In fact some of the testing done here was with drives just laying outside of an open computer case! This LGA 1156 system will be repurposed; it will loose some of its parts, have its CPU set back to stock speed (which is the way its always has been until one month ago), and it will be used in our family room as a "Home Theater PC."

             

            I'll be really curious to see if my new LGA 1366 (X58 chipset) build with a i7-950 CPU will perform closer to Harm's record setting quad core score [PPBM5 composite 158 secs.] or closer to this LGA 1156 (P55) system [PPBM5 composite 195 secs.]. Anyone want to venture a guess?

             

            Thanks again to you and Harm for all of your help!

             

            Jim

            • 3. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
              Harm Millaard Level 7

              Jim,

               

              With all the testing you have done and the experience you gained in the process, I think it is not overly optimistic that you can mimic Cristobal's results, maybe even improve on them (182 s), depending of course on your configuration. Whether you can beat my score is largely determined by the amount of memory and your disk setup. I know, many consider my disk setup excessive.

              • 4. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                Jim_Simon Level 8

                One item I noticed missing was how the drive testing affected the MPEG2 and H.264 portions of the benchmark.

                 

                I personally find that including disk I/O in the "overall score" as somewhat befuddling.  The actions of data copying and data processing are very different.  Having both results included in an overall score can skew perceptions.  Someone with a very fast disk I/O could potentially have a very low overall score, but have slower encoding times vs. someone with a fast CPU and a slow disk I/O.  It can be difficult to read at a glance that the lower scored rig is actually a worse performer when it comes to choosing the right CPU to buy.

                 

                I'd love to see the disk and processing sections kept separate in the benchmark.  They're not likely to affect each other significantly, but combining the results can distort the at-a-glance picture of the system.

                • 5. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                  Bill Gehrke Adobe Community Professional & MVP

                  Jim, I have to ask have you seen what Harm has added to the home page of PPBM5?  The four seperate sortings allow you to sort the todays total of 233 results (2/16/11) by for instance Disk I/O scores to compare the disk intensive results.

                   

                  At some point I will be publishing data that I have been gathering of total performance versus disk systems.

                  • 6. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                    JEShort01 Level 4
                    function(){return A.apply(null,[this].concat($A(arguments)))}

                    Bill Gehrke wrote:

                     

                    Jim, I have to ask have you seen what Harm has added to the home page of PPBM5?  The four seperate sortings allow you to sort the todays total of 233 results (2/16/11) by for instance Disk I/O scores to compare the disk intensive results.


                    Bill,

                     

                    I have indeed seen and used the PPBM5 "sorted" results (off of the menu); in fact, I was copying and pasting the PPBM5 results into Excel for months so that I could "study" various performance metrics and while focusing on the results from produced by quad-core CPUs (i.e. ignoring the $1000 and up per CPU chip models).

                     

                    As I noted in my results from testing above, I found it interesting and a bit curious that the PPBM5 I/O results were impacted significantly by changing the clock speed on the CPU. On my own, I did some testing on different drive array configurations where I copied a 25GB file around, which to me is a more pure Disk I/O test.

                     

                    Thanks,

                     

                    Jim

                    • 7. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                      JEShort01 Level 4
                      function(){return A.apply(null,[this].concat($A(arguments)))}

                      JSS1138 wrote:

                       

                      One item I noticed missing was how the drive testing affected the MPEG2 and H.264 portions of the benchmark.

                       

                      I personally find that including disk I/O in the "overall score" as somewhat befuddling.  The actions of data copying and data processing are very different.  Having both results included in an overall score can skew perceptions.  Someone with a very fast disk I/O could potentially have a very low overall score, but have slower encoding times vs. someone with a fast CPU and a slow disk I/O.  It can be difficult to read at a glance that the lower scored rig is actually a worse performer when it comes to choosing the right CPU to buy.

                       

                      I'd love to see the disk and processing sections kept separate in the benchmark.  They're not likely to affect each other significantly, but combining the results can distort the at-a-glance picture of the system.

                       

                      For MPEG2-DVD and H.264 BR tests, here are some additional specifics:

                      SetupMPEG2-DVD (secs)H.264 BR (secs)
                      CPU at 2.8 GHz (SSD boot, all files on RAID5)48107
                      CPU at 2.93 GHz (SSD boot, all files on RAID5)47103
                      CPU at 3.08 GHz (SSD boot, all files on RAID5)4597
                      CPU at 4.0 GHz (SSD boot, all files on RAID5)3781
                      CPU at 4.0 GHz (all other runs w/ 16GB RAM)from 37 to 38from 78 to 81
                      reduced RAM to 8GB9183
                      reduced RAM to 4GB12289

                       

                      As you can see, and as I previously summarized, CPU speed does affect all 4 core components of the PPBM5 test and the RAM size is very important for the MPEG2-DVD test.

                       

                      Hope this helps!

                       

                      Jim

                      • 8. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                        Bill Gehrke Adobe Community Professional & MVP

                        My favorite Disk I/O test is not very convienent but I have on occasion set up a RAM Disk taking 10 of my 24 GB and then using half of the PPBM5 AVI raw file (6.63 GB) for my testing.  If you want to see some speed look at these HD Tune Pro results below.  Now the problem is that you need almost 14 GB of space to run PPBM5 and if I were to make a 14 GB RAM Disk that would cripple the results running with only 10 GB of RAM--if only those 8GB sticks were available!  But it was great having a very fast source to find the real WRITE speed of my large RAID 0 array, and having a destination (RAM) disk the allows getting the true READ speed of the array.  Unfortunately timing the operation is a major problem.  I had to use a stop watch and real accuracy was a problem.

                        HDTune-RAMDisk-Read.jpg

                        HDTune-RAMDisk-Write.jpg

                        • 9. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                          JEShort

                          Bill,

                           

                          That is smokin' fast! What software do you use / recommend for your RAM disk?

                          • 10. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                            Jim_Simon Level 8
                            Hope this helps!

                             

                            Actually, I was hoping to see how the changes to disk setup affected the MPEG and H.264 portions.  I get that the CPU will have an impact.

                            • 11. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                              JEShort01 Level 4
                              function(){return A.apply(null,[this].concat($A(arguments)))}

                                Actually, I was hoping to see how the changes to disk setup affected the MPEG and H.264 portions.  I get that the CPU will have an impact.


                              Jim,

                               

                              Bottom line was that the disk setup did NOT affect MPEG and H.264 portions hardly at all; for example for the MPEG test with project files, media cache and media cache DB all on a single 1TB 7200 rpm hard drive the result was 38 seconds and with the project files on an 8x1TB RAID 5 array and media cache and media cache DB on a 2xSSD RAID0 the result was 37 seconds.

                               

                              Similarly, the H.264 result using the single 1TB 7200 rpm drive was 81 seconds, and the fastest drive setup tested brought that time down to 78 seconds.

                               

                              Probably you were expecting more of a difference, I know that I was, but it just was not there.

                               

                              Jim

                              • 12. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                Jim_Simon Level 8
                                Probably you were expecting more of a difference

                                 

                                Actually, no.  I expect very little difference, which is why I think the "overall score" could be misleading.

                                 

                                The disk and processing portions should be kept fully separate.  (Harm? Bill?)

                                • 13. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                  Harm Millaard Level 7

                                  Jim,

                                   

                                  Do you agree the four basic ingredients of an editing rig are:

                                   

                                  ComponentMainly Tested with:
                                  Performance Weight
                                  Main Discriminating Factor
                                  Average Score
                                  CPUH.264-BR25%Number of cores / L3 cache110
                                  MemoryMPEG2-DVD25%Amount of memory101
                                  VideoMPE Render25%CUDA card or not13-MPE on / 126-MPE off
                                  DiskAVI export25%Disk speed135

                                   

                                   

                                  If one of these four is missing, you have no editing rig. You need all four. And these components are emphisized in the 4 tests as shown above.

                                   

                                  System performance is determined by the weakest link in the chain. Each test has a 25% weight in the total score, equal weights for each of the basic components. One can dispute the arbitrary weight of 25%. If you never render your timeline, you should attach less weight to the MPE render test. If you make it a custom to never finish a project, export is not really relevant and disk speed is not very relevant. If you never encode to MPEG2-DVD, you can attach a smaller weight. If you never edit AVCHD, because you hate the codec, the weight for H.264-BR can be lowered.

                                   

                                  Those considerations vary so much that each individual must interpret the results with that in mind. It would be a great disservice to leave out one test, and disregard that an editing rig should have all four basic components and that overall performance depends on the weakest link.

                                   

                                  What do you mean with:

                                   

                                  The disk and processing portions should be kept fully separate.

                                   

                                  No disk results and no CPU results and no memory results? How do you want to leave out the CPU, memory and disk, because you can't run video without the CPU, memory and disk.

                                   

                                  I have the feeling you did not really look at the site, the various pages and the different results pages with anything more than a cursory glance. If you did look at all the material and you misunderstood, it must be my lacking English, because I wrote the text. In that case, maybe you can help me edit the text, so it is better understood.

                                  • 14. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                    Crist OC/PC Level 2

                                    Ussing short words. The Relevant test is CPU / GPU MPE OFF.  you can compare this whit the best result, them decide how to impruve your system.

                                     

                                    Once again congrats To harm and bill. Fantastic job with the ppbm.

                                    • 15. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                      Bill Gehrke Adobe Community Professional & MVP

                                      Darn forum I did not believe the first one went through.

                                      • 16. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                        Bill Gehrke Adobe Community Professional & MVP

                                        Crist OC/PC wrote:

                                         

                                        Ussing short words. The Relevant test is CPU / GPU MPE OFF.  you can compare this whit the best result, them decide how to impruve your system.

                                         

                                        Once again congrats To harm and bill. Fantastic job with the ppbm.

                                        I think I have to disagree.  There may be a good level of correlation on the first few top entries be as you go down the list that theory falls apart.  That is precisely why Harm came up with the performance index.

                                        • 17. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                          Jim_Simon Level 8
                                          It would be a great disservice to leave out one test

                                           

                                          I agree.  But I wasn't suggesting any of the four tests be left out of the benchmark.  I was suggesting that the disk speed results be left out of the "overall score", as they do not help with providing a picture of how fast a system may be with CPU intensive tasks.  Quite the opposite, in fact.  Including them can skew perceptions about how one rig performs against another.

                                           

                                          Keep all tests, but keep the disk results and CPU results separate in the scores.  This gives a more accurate "at-a-glance" picture of how rigs compare.

                                           

                                          And I'd recommend ordering the results with the render scores, sans disk scores.  Leave those as as 'aside'.  This will provide a much more accurate picture of how rigs compare with processing intensive tasks, and allows one to make a better CPU purchasing decision.

                                           

                                          Does that clarify my point sufficiently?

                                          • 18. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                            Harm Millaard Level 7

                                            Not really Jim. Let me try to explain.

                                             

                                            There are essentially three distinct phases in editing:

                                             

                                            1. Ingest, which includes writing source material to disk, indexing, conforming and creating peak files and writing to the media cache and media cache database.

                                             

                                            2. Editing, which includes reading source files and media cache files from disk, writing and reading preview files and auto save files.

                                             

                                            3. Export, which includes reading the timeline material and writing transport or elementary streams back to disk.

                                             

                                            None of the three essential phases can be done without disks. BTW, that is exactly why both MPEG2-DVD and H.264-BR tests show a very small impact of disk setup, because with the short timelines, there is only a small amount of data to be written to disk. 98 MB and 155 MB respectively. In practical life, one would export much larger files, say 4.0+ GB for MPEG2-DVD and 20+ GB for H.264-BR. When that happens, the write speed is crucial, but that can not be expressed with this benchmark. Only the writing of a 13 GB file with the AVI test does show how disk setup impacts on export times.

                                             

                                            Let's have a look at two significant results and take note of the colors and legends in the scores of Maxpc and GreenFX_5:

                                             

                                            19-2-2011 12-02-13.jpg

                                             

                                            As you can see at a glance, Maxpc had a very decent disk setup, with a score of 115, but his CPU intensive results are lousy, all near the bottom of the list. GreenFX_5 has lousy disk results, but his CPU intensive results are very good.

                                             

                                            Maxpc is hampered by the limitations of the Phenom II X6. Nothing he can do about that, apart from adding additional memory and he has made sure that his systems performance is not degraded by disk performance. GreenFX_5 however has a very fast CPU, but in real life his export times will be extremly bad, because his disk performance is an incredible bottleneck. Practical export results will at best be mediocre in comparison to much lesser specced systems, as shown by the RPI. His 980X is not using its full potential because of the imbalance in his components.

                                             

                                            These two examples are extreme, but show that your desire for seeing 'at a glance' what is going on is reflected in the colors and legends used.

                                             

                                            Your statement:

                                             

                                            allows one to make a better CPU purchasing decision.

                                             

                                            is too restrictive IMO. It is all about a well balanced system, not only the CPU. I repeat that a system only performs as well as its weakest link allows, not the best, as is clearly demonstrated by the examples above. In some cases one can not do much about the weakest link without a complete overhaul of the system (Maxpc), but in other cases (GreenFX_5) improvements are relatively easy.

                                             

                                            I sincerely believe that with our RPI, Relative Performance Index, and the color coding and legends in the results columns, and the ability to sort all results on each test, we have done users a better service than by following your suggestion of taking out the disk results out of the total score. GreenFX_5 would be delighted by his improved ranking, but it would not warn him of the serious shortcoming in his system setup.

                                             

                                            Hope this explains a bit why we did it this way.

                                            • 19. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                              Crist OC/PC Level 2

                                              Bill, what could happen if Harm rise CPU up to 4,2 GHZ?

                                               

                                              I guess CPU test H.264 BR, Disk I/O test, CPU / GPU MPE OFF will go down. Giving him a better time. No doubt  he has the BEST disk setting in the PPBM. and that make the difference whit others I7920,30 or 50. And that is just a good indicator of how much we can improvise in  Disk I/O test.

                                               

                                              in other hand part of GPU MPE OFF depends on your HDISK config.

                                               

                                              Please corrects if  I am mistaken, as this is very intersting for my.

                                               

                                              Thx.

                                              • 20. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                Bill Gehrke Adobe Community Professional & MVP

                                                JEShort wrote:

                                                 

                                                Bill,

                                                 

                                                That is smokin' fast! What software do you use / recommend for your RAM disk?

                                                I used the DataRAM RAMdisk software for those results.  I do not have it running normally.  If I remember correctly the unlimited (RAM) version was $15.  There is a 4 GB version that is freeware.

                                                • 21. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                  Bill Gehrke Adobe Community Professional & MVP

                                                  All the five PPBM5 tests have some CPU speed/cores dependencies.  If you notice for instance even when running our disk intensive AVI encoding test the CPU usage is around 30 percent.(on an i7-980X)

                                                  • 22. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                    JEShort Level 1
                                                    function(){return A.apply(null,[this].concat($A(arguments)))}

                                                    Harm Millaard wrote:

                                                     

                                                    Jim,

                                                     

                                                    With all the testing you have done and the experience you gained in the process, I think it is not overly optimistic that you can mimic Cristobal's results, maybe even improve on them (182 s), depending of course on your configuration. Whether you can beat my score is largely determined by the amount of memory and your disk setup. I know, many consider my disk setup excessive.

                                                     

                                                    Harm,

                                                     

                                                    As you predicted, the new X58 build is faster on PPBM5 than the i7-860, but only by about 10% (170 seconds vs. 195 seconds).

                                                     

                                                    I'll post my PPBM5 score later today as JES2:170 seconds (Disk I/O 69, MPEG-DVD 31, H.264 BR 66, TL MPE 4, and TL no MPE 79).

                                                     

                                                    In your opinion what would be the best way to add and configure 5 additional drives to this system (full listing below) from my current "spares":

                                                    - (5) older 1TB Hitachi (7k1000.c; 7200rpm)

                                                    - (3) 1TB WD Black/RE3 drives (32MB versions; 72000 rpm)

                                                     

                                                    Current configuration is:

                                                    CPU i7-950 (OC'd to 4.2GHz) water cooled and silent

                                                    MB Gigabyte GA-X58A-UD7

                                                    GPU MSI GTX 480 Hydrogen (OC'd to 900MHz, memory at 2000MHz) water cooled and silent

                                                    Memory: 24GB (6x4GB G. Skill 1600 CAS9)

                                                    Case / power supply: Coolermaster HAF 932 Black Edition (no side fan - quiet) / Corsair AX1200 (big fan, also quiet)

                                                    OS Microsoft Windows 7 Ultimate 64-bit

                                                    OS drives:5x64GB Raid 0 Crucial C300 SSD (currently used for Boot OS, programs, Premiere Pro media and media cache DB)

                                                    Project drives on Areca ARC-1660ix-12 (2MB cache): 5x1TB WD Black/RE3 mix Raid 0 (7 channels currently unused)

                                                     

                                                    Also, how do you configure PPBM5 files, cache, cache DB, scratch, etc. for your best times? Do you split up the various files or simply put everything on your monster R30 to gain your best results?

                                                     

                                                    Thanks,

                                                     

                                                    Jim

                                                    • 23. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                      Harm Millaard Level 7

                                                      Jim,

                                                       

                                                      I just recieved you form, but noticed you have not yet updated to the new script. Nothing for you to do about it at the moment, I will handle it, but if you want to resubmit new data in the future, I appreciate it if you were to use the updated script.

                                                       

                                                      I'll get back to you on the other questions, but I'm currently on-line with someone in Egypt to help him out.

                                                      • 24. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                        JEShort01 Level 4
                                                        function(){return A.apply(null,[this].concat($A(arguments)))}

                                                        Harm Millaard wrote:

                                                         

                                                        Jim,

                                                         

                                                        I just recieved you form, but noticed you have not yet updated to the new script. Nothing for you to do about it at the moment, I will handle it, but if you want to resubmit new data in the future, I appreciate it if you were to use the updated script.

                                                         

                                                        I'll get back to you on the other questions, but I'm currently on-line with someone in Egypt to help him out.

                                                         

                                                        Harm,

                                                         

                                                        Thanks, the updated script is much faster to use as well!

                                                         

                                                        Regarding, "I'll get back to you", any thoughts?

                                                         

                                                        Thanks,

                                                         

                                                        Jim

                                                        • 25. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                          Harm Millaard Level 7

                                                          Sorry Jim for not getting back to you sooner, but I had some tuning to do on the new results pages, that took more time than I anticipated. Crappy Excel exports to .HTML with conditional formats in the spreadsheet.... It is a known bug, but it does cause a lot of grief.

                                                           

                                                          I have my media cache and media cache database files on the same 2 disk raid0 array as my pagefile. All the rest is on my 12 disk raid30 array.

                                                           

                                                          With your current disks, I would use the 3 WD's in a raid0 and the 5 Hitachi's in a raid3/5 and then the separate OS. You could use the same setup I used, the raid0 for media cache, database and pagefile and the rest on the raid3/5. I would then extend the raid3/5 to a larger number of disks in the future.

                                                          • 26. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                            JEShort01 Level 4
                                                            function(){return A.apply(null,[this].concat($A(arguments)))}

                                                            Bill Gehrke wrote:

                                                             

                                                            Jim, great summary. This is one very good example of one of the prime purposes of this benchmark, to optimize what ever you have.

                                                             

                                                            Congratulations you have the highest scoring LGA 1156 board based system on the test results.  On your submitted data in the PPBM5 Results page all your results fall into the Q3 catagory which indicates you have a very well balanced configuration.

                                                             

                                                            Now you can see why we have been recommending the even slightly faster X58 class processors, more than 12 GB of RAM, motherboards (LGA1366) and do not generally recommend SSD drives.

                                                             

                                                            Thank you.

                                                             

                                                            Bill

                                                             

                                                            Harm and Bill,

                                                             

                                                            Harm's top PPBM5 quad-core score was tough to top, but I've done so using SSD and hard drive arrays where each is strongest. I just uploaded the PPBM5 files for JES3 with a composite time of 156 seconds with an overclocked i7-950:

                                                             

                                                            - OS and project files on a 5x64GB Crucial SSD RAID0 array

                                                            - media cache, media cache DB, and output files on a 12x1TB RAID0 array (PPBM5 submission may incorrectly say 10x?)

                                                            - times for each component of PPBM5: Disk I/O 66 seconds, MPEG2 22 seconds, Blu-Ray 62 seconds, timeline w/ MPE 4 seconds, timeline w/o MPE 78 seconds (composite score 156 seconds)

                                                             

                                                            Other thoughts:

                                                            - PPBM5 score of 156 seconds using a $199 CPU (Micro Center in store pricing for i7-950 in the USA) = high bang for the buck - hardly, there is so much more to a decent Premiere Pro CS5 system and it all adds up!

                                                            - PPBM5 is a great analysis tool - thanks again Harm and Bill!

                                                            - this build is NOT balanced, and 20 drives is indeed excessive for a home PC (20 drive count includes 3 1TB drives currently connected for backups)

                                                            - I will be doing additional testing on other applications and scaling back to something less excessive (some excessives will remain I'm sure!) in the drive area

                                                            - I have run PPBM over 100 times now - whew! Time to return back to my life...

                                                             

                                                            Regards,

                                                             

                                                            Jim

                                                            • 27. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                              Harm Millaard Level 7

                                                              Jim,

                                                               

                                                              Congrats!. I haven't yet seen your results (I'm on my notebook now), but you are right, the submission form does not allow for more than 10 disks for a single level array, because we think that no editing rig would have more than 10 disks without some security built in. Nobody would run a 10 disk raid0 on an editing rig, let alone 12 disks. But you still have the comments field for that.

                                                              • 28. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                                Crist OC/PC Level 2

                                                                Really well done Jim. Thx for all your imputs and coments. Congrats.

                                                                • 29. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                                  JEShort01 Level 4

                                                                  After a cpu upgrade and some more tweaking...

                                                                   

                                                                  PPBM5 = 134 seconds (composite)

                                                                  Disk I/O = 59 seconds

                                                                  MPEG2-DVD = 18 seconds

                                                                  Blu-ray = 53 seconds

                                                                  Timeline = 4 seconds

                                                                   

                                                                   

                                                                  Here's what has changed:

                                                                  CPU - Upgraded to i7-970 (from i7-950)

                                                                  Drive arrays:

                                                                  - Boot (5x64GB Corsair SSD RAID 0) = OS, Win7 swap, media cache DB

                                                                  - Data (7x1TB WD Black/RE3 mix RAID 0) = projects, media, scratch

                                                                  - Output (5x1TB WD Black RAID 0) = media cache, output

                                                                   

                                                                  Regards,

                                                                   

                                                                  Jim

                                                                  • 30. Re: How changes to CPU speed, RAM size, and drives (SSD and hard drives) impacted my PPBM5 score
                                                                    Crist OC/PC Level 2

                                                                    sweet =). Congrats Jim, I´m looking forware to upgrate to 8 cores and hi OC.

                                                                    One more time, well done.

                                                                    B.R.

                                                                    Cristobal Salas