15 Replies Latest reply on Feb 27, 2013 3:42 PM by hpmoon

    Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW

    Luke Humphrey

      I've been wanting to make some major upgrades to my hardware, but it just doesn't seem worth it yet...even after almost 4 years. I ultimately decided to "rent" a new video card and run some tests. Here is some background info on my upgrade though process and some results comparing the video card performance.

       

      Disclaimer: I'm not a hardware expert, but I'm not completely clueless (I think). Your input/insight is welcome.

       

      My system (purchasd 2/2009)

       

      i7 920

      GTS 450 (1GB RAM)

      12 GB 1333 RAM

      Samsung SATA II 128 GB SSD (OS/apps)

      5x 1TB 7200 RPM drives in RAID 0 (with accompany slow/cheaper 2TB backup drives)

       

      Some upgrade options I am considering

       

      Sandy Bridge 3930 - but it's $560 w/o cooling and would require a new, more expensive motherboard, new ram, cooling, etc.

      Ivy Bridge 3770, but I keep reading that that an overclocked 920 isn't that much different in perf (in fairness mine isn't oc'd). I did find a MB that would work for only $90. So I could make this upgrade for just under $400 (RAM would stay the same).

      Wait for Haswell, but i could be another 9 months and it's supposed to only give maybe a 10% perf gain over IB. It's more focused on mobile - less power, integrated graphics, etc.

      High-end Xeons are totally off the table. $/buck is waaaay too low.

       

      Video card and benchmark reviews/problems

       

      So I thought I'd first try getting a new video card. I see conflicting benchmarks. This site (the one that provides the CUDA.exe hack) notices very little difference between most GTX cards in perf for their benchmarks. The PPMB5 site shows significant differences between say the GTX 680 and lower end cards. But are these really accurate?

       

      The GTX 680 is almost $500, so I opted for the 660 ti at $300 to see if I could get a noticable perf gain. It seemed like the best $/buck card and wouldn't require me to get a new power supply.

       

      Another reason I wanted to do my own tests: None of the benchmarks I've seen actually mention the type of footage used. I care about footage from the Canon MKII-III, or similar footage. I definitely do not care about things like exporting to MPEG 2.

       

      I did some very unscientific benchmarks, but they were real world for me. First my "problem" areas.

       

      Performance problem areas

       

      #1 - Time-lapses consisting of 1080 (height) JPEGs and 2160 (height) JPEGs don't always play smoothly (larger 2160s almost never do). I read adding more VRAM might help. The 660ti has 2x the RAM as my current video card.

      #2 - Split screen sequences (up to 9 clips simultaneously) don't play smoothly.

      #3 - Scenes where I speed up a clip to 1000x don't always play smoothly. (Although upgrading from CS5 to 6 actually seems to have solved this issue, I couldn't get it to repro any longer).

      #4 - Export to h.264 could be faster. I do this a lot, but mostly because it's how I sometimes make proxies because of problems around #1-2 (works fine - used to use CineForm but it always crashed Premiere and these work for my needs). This is typically my final export as well for posting on sites like Vimeo.

      #5 - Timeline rendering could be faster, although I don't do this a lot and if I do it's simple, not a bunch of crazy effects. E.g. use unsharp mask. This is pretty low pri for me though because I think timeline rendering is a bad idea. Once you do it, if you even move the clip you have to render again.

       

      Some simple bottleneck analysis first:

       

      Disk queue length sometimes is just over 1 on 1 disk in my RAID array during TL playback. Might slow things down slightly. Not an issue during export.

      Processor never seems to get pegged in any case.

      RAM is never maxed out, but it starts to go to Premiere limits (10 GB that I've set) after playing through several time-lapses (I'm just now noticing this). Choppiness starts well before RAM is even near that on some clips.

       

      Tests/results:

      NOTE: I do run the 660ti in a PCIe 2 x16 slot. Let me know if you think it would even matter to run in a PCIe 3.0 slot. My MB doesn't have one.


      #1 Time-lapse smoothness - didn't improve with the 660. Moving the 1080 size JPEG TLs to my SSD did help some problem TLs play smoothly however.

      #2 Split screen. Did a test with a 9-clip-at-the-same-time sequence. No improvement with the 660ti.

      #3 Clips speed up 1000x - could not repro the problem now that I run CS 6 vs. 5 on either card.

      #4 - Export to H.264 1080p @23.9x fps.

       

      Export 5:30 clip of 5D MKIII footage + H.264 proxies:

       

      GTS 450 - 9:14

      660 Ti - 8:30

       

       

      Export 1.5 minute clip of large time-lapses (JPEGs that are 2160 high):

      GTS 450 - 9:35

      660 Ti - 7:00

       

      Export a 2 minute clip of just MKIII footage

       

      GTS 450 - 2:45

      660 Ti - 2:45

       

       

      #5 Timeline render with simple image correction effect

       

      Timeline render short 5D MKIII clip with unsharp mask applied:

      GTS 450 - 1:10

      660 Ti - 1:19

       

      Conclusion:

      The 660ti ($300) showed marginal improvements in exporting h.264 against my GTS 450 ($100) and did not address my other issues. Definitely not worth it for the type of work I do.

      Moving my time-lapse JPEGs to an SSD helps play the 1080p versions back smoothly. The 2160p larger versions still lag. Maybe more RAM would help? They still start off choppy and then acquire more and more RAM, so not sure here. Maybe faster 1600 RAM? I don't know, I doubt it. I may have to just use 1080 versions or make proxies.

      I don't see a pegged CPU much if at all, so upgrading to an Ivy Bridge 3770 doesn't seem like it'll help much if at all.

      I did end up buying 2x256 GB SATA III SSDs (only $169 each) that I'll run current projects off of, or at least time-lapse sequences (RAID 0). My motherboard doesn't have an SATA III slots, however, so I won't see the full power of these, but not sure I'll need it. Again I'm not seeing a clear disk issue either from the perf monitoring.

       

      I suspect many of these problems are still with the software and how it takes advantage of my hardware, but I'd love more insight.

       

      Generally I make things work and I don't have any really painful bottlenecks, but I'm always up for perf improvements/doing things faster. It does look like I won't see any major breakthroughs, however by spending $400-$1000 bucks on HW upgrades.

       

       

      Thoughts?

       

      Luke

      Blog  |  Photography  |  Vimeo

        • 1. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
          Harm Millaard Level 7

          Luke,

           

          A number of remarks about your system, video card results, testing results, etcetera.

           

          SYSTEM: It is an older system, about the same I had in the form of 'Harm's Beast', although I have a much beefier disk setup, more memory and OC'ed to 3.7 GHz, in combination with a GTX 480. Not much you can do about this system, apart from upgrading memory to 24 GB but the major drawback is that those investments will not carry over to a new system, at least not easily.

           

          VIDEO: You correctly point out that the GTX 680 shows in the MPE graph on the PPBM5 website much better results than other cards. But keep in mind that most 680's are used in new systems, often with the latest CPU's and fast memory. I'm convinced that a 680 is not noticeably faster than a 580, because they have the same memory bandwidth, but it looks that way because they are often accompanied by hexa core i7-39xx CPU's with large amounts of memory.

           

          The GTS 450 has a memory bandwidth of 86 GB/s, the 660 Ti has 144.2 GB/s, so the latter is significantly faster as you have shown in some of your tests.

           

          TESTING: You don't mention to what format you exported and with what resolution and framerate. Hardware MPE will come into play when you have rescaling, frame blending, blurring and stuff like that occurring. If you export to the same frame size and frame rate as your source and no blurring occurs, then exporting is purely a CPU matter and the video card has no impact at all.

           

          General remarks: I personally consider your 5 disk raid0 setup as pretty risky. You have multiplied the risk of losing all data by a factor 5!. You have no redundancy at all. Even though it is fast, I expect your sustained transfer rates are less than 450 MB/s and when using a 9 clip split screen, it may be too slow with the limited memory and the old CPU you have.

           

          You have effectively one single volume for video related editing (apart from the OS disk) and while that makes for easy administration, it still entails the drawbacks of the half-duplex connection of SATA. It might be better to add a couple of HDD's in raid0 for media cache, previews and exports to avoid that limitation. You can always carry those to a new system.

           

          Sorry to be so harsh.

          • 2. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
            Luke Humphrey Level 1

            Thanks for the resonpose Harm. Inline.

            Harm Millaard wrote:

             

            SYSTEM: It is an older system, about the same I had in the form of 'Harm's Beast', although I have a much beefier disk setup, more memory and OC'ed to 3.7 GHz, in combination with a GTX 480. Not much you can do about this system, apart from upgrading memory to 24 GB but the major drawback is that those investments will not carry over to a new system, at least not easily.

             

            [Luke] From your description of your system it sounds like 4 things could indeed be upgraded and carried over to a new system. 1) OC the processor (e.g. purchase a generic water cooler for ~$100), 2) Improve the Disk setup, 3) Upgrade the video card, 4) Add more/faster RAM.

             


            I've seen in some benchmarks that an OC'd 920 is not so dissimilar to an OC'd 3770K. The latter is faster, but it isn't a huge difference. The larger question still remains - will any/all of these upgrades yield large performance gains and solve all/a higher percentage of my problems? Or do I have a decent sweet spot of a system and should wait for the software (e.g. MPE evolution in CS7-8) to catch up and take better advantage of what I have? Like I said from doing some rudimentary performance monitoring, I'm not seeing a pegged CPU (just a brief spike here/there), I'm not seeing disk transfer at capacity (although 1 disk has a slightly > 1 queue length at times), I'm not seeing in all cases over-utilization of memory, etc. (except higher RAM usage is seen albeit staggered for large JPEG time-lapse sequences, but I see choppiness well before RAM usage gets to 10 GB).


             

            VIDEO: You correctly point out that the GTX 680 shows in the MPE graph on the PPBM5 website much better results than other cards. But keep in mind that most 680's are used in new systems, often with the latest CPU's and fast memory. I'm convinced that a 680 is not noticeably faster than a 580, because they have the same memory bandwidth, but it looks that way because they are often accompanied by hexa core i7-39xx CPU's with large amounts of memory.

             

            [Luke] Good point - potentially further evidence that the video card doesn't make a big difference? At least not enought to justify 5x the cost (e.g. $500 680 vs. $100 450).  This would be consistnet with what Studio 1 Productions has seen.

             

            The GTS 450 has a memory bandwidth of 86 GB/s, the 660 Ti has 144.2 GB/s, so the latter is significantly faster as you have shown in some of your tests.

             

            [Luke] The only test I would characterize closer to having a significant increase using the 660 would be exporting large JPEG time-lapses to H.264, where it was a good 27% faster. The rest seemed more marginal or did not change.

             

            TESTING: You don't mention to what format you exported and with what resolution and framerate. Hardware MPE will come into play when you have rescaling, frame blending, blurring and stuff like that occurring. If you export to the same frame size and frame rate as your source and no blurring occurs, then exporting is purely a CPU matter and the video card has no impact at all.

             

            [Luke] See above - H.264, 1080p @23.9x fps.

             

            General remarks: I personally consider your 5 disk raid0 setup as pretty risky. You have multiplied the risk of losing all data by a factor 5!. You have no redundancy at all. Even though it is fast, I expect your sustained transfer rates are less than 450 MB/s and when using a 9 clip split screen, it may be too slow with the limited memory and the old CPU you have.

             

            You have effectively one single volume for video related editing (apart from the OS disk) and while that makes for easy administration, it still entails the drawbacks of the half-duplex connection of SATA. It might be better to add a couple of HDD's in raid0 for media cache, previews and exports to avoid that limitation. You can always carry those to a new system.

             

            [Luke] Yes there is a higher level of risk, but with backups every 30 minutes during project work I yield cheap/easy perf gains for the cost of--at most--30 minutes of work. I've lost no work in the last 4 years, 1 drive failed once while on vacation and I replaced it easily. Anyway backup/data integrity is a different issue separate from performance which I'd like to focus on in this context.

             


            I get ~420MB/s read with this array (mostly older, blue WD drives and Deskstars). I'm running out of space, so I just ordered 3 x 2 TB WD black drives to replace this with, expecting probably a similar transfer rate. Again though I'm not necessarily seeing disk being a bottleneck in perf mon aside from one disk who's queue length sometimes goes over 1, so we'll see if the newer black drives help.

             

             

             


            I have ordered 2x256 SATA III SSDs to put min my time-lapses on as having them on my current primary SSD seemed to help in some cases.

             

            Sorry to be so harsh.

             

            [Luke] No worries, harsh is OK but I'm still not seeing a clear solution to some of my issues and I'm still not convinced a new system - short of a top-of-the-line SB or Xeon system (both of which are very $$) will be worth the upgrade. When 5 years had passed between my current system and my 2004 system, I feel like the upgrades were much more significant especially for bang/$.


             

            --

            Luke Humphrey

            Blog | Photography | Cinematography

            • 3. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
              Luke400 Level 1

              Update: I upgraded my system: Z77 motherboard, 3770 IB processor, 32 GB 1600 ram, 2xSATA III SSDs RAID 0 (read speeds over 1GB/s). Kept the Nvidia GTS 450 and took back the 660 ti.

               

              Results with upgraded disk, RAM, processor (3770):

               

              #1 - Time-lapse smoothness (4000x2667 jpgs)

              Old system (GTS 450) - Choppy

               

              Old system (660 ti) - Choppy

              New system (GTS 450) - Little to no improvement , even with 1GB/s read speeds on the new SSDs. CPU runs at 55%, Ram got up to 23 GB by the end of the clip.

               

               

              #2 - 9 different clips of 5D MKII footage tiled (played simultaneously) - timeline playback

               

              Old system (GTS 450) - Choppy

               

              Old system (660 ti) - Choppy

              New system (GTS 450) - Smooth playback! (repeated several times, all smooth)

               

               

              #4 - Export to H.264 1080p @23.9x fps (most footage stored on 3x2TB 7200 RPM RAID array, read speeds ~500MB/s)

               

              4a - Export 5:30 clip of 5D MKIII footage + H.264 proxies:

              Old system (GTS 450) - 9:14

              Old system (660 ti) - 8:30

              New system (GTS 450) - 6:10

               

              (NOTE: Ram anywhere from 7-13 GB, CPU mostly at 70-90% but sometimes at ~100%)

               

              4b - Export 1.5 minute clip of large time-lapses (~4000x2667 jpgs):

              Old system (GTS 450): 9:35

              Old system (660ti): 7:00

              New system (GTS 450): 3:52

               

              (NOTE: Ram hovers around 16-17 GB, CPU mostly at 80% but sometimes at ~100%)

               

              4c - Export a 2 minute clip of just MKIII footage

              Old system (GTS 450): 2:45

              Old system (660ti): 2:45

              New system (GTS 450): 2:17

               

              #5 Timeline render of MKIII footage with simple image correction effect

               

              Old system (GTS 450): 1:10

              Old system (660ti): 1:19

              New system (GTS 450): 0:45

               

              Conclusion

              The thing I care about most was the large time-lapse smooth playback, and that did not improve unfortunately. Maybe 2800 speed ram would help? I don't know. It's certainly not a disk issue as I had the JPGs on the 1 GB/s raided SSDs.

               

              Overall a huge improvement in exporting to H.264, timeline rendering a non-GPU accelerated basic effect, and buttery smooth showing 9 MKII clips at the same time tiled.

               

              If you have any thoughts on improving the large time-lapse playback smoothness, please let me know.  I'm going to try exporting lower quality versions to see if that helps. I suppose, however if I'm going to do that I might as well just scale up the 1080p versions that do playback smoothly. Will experiement - I was hoping for a hardware-based solution that would allow for full quality during timeline playback.

               

               

              Luke Humphrey

              Blog | Photography | Cinematography

              • 4. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                Luke400 Level 1

                OK I found a solution to the 2.5k timelapse smooth playback problem. If I export at 3840x2560 from Lightroom at JPEG quality 60 to the SSDs, I retain high-quality for timeline editing and that seems to be a sweet spot for smooth playback on this hardware (or maybe it's quality 70, need to play more). I need to test it with a larger variety of time-lapse clips, but so far so good at these settings. I was able to zoom in and speed up one clip by 300% and retain smooth playback.

                 

                At this modified size/quality setting - It works out to just under 1MB image vs. 6-7MB/image with higher quality settings, so at 24fps that's 24MB/s vs. 168 MB/s. Why 168 MB/s is too much for a disk system that reads up to 1 GB/s I'm not sure.

                 

                 

                Side note: I was also hoping my upgrade would improve Lightroom perf. Whenever I cycle through images there's always a pause and a slight "loading" while the CPU is pegged. No real change with my rig upgrade, I'm guessing there is a software optimization issue there.

                • 7. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                  RjL190365 Level 4

                  Thanks, Luke, for those results. It should be noted, however, that some of the improvement in performance came from the extra 20GB of system RAM in the new system (your old system had only 12GB of RAM, which IMHO is not enough to run Premiere Pro CS6 to an enthusiast's satisfaction since in my testing CS6 needs 24GB or more for optimal performance).

                  • 8. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                    Luke400 Level 1

                    Much of the time the RAM was around 8-10GB utilitization, but yes I saw a big jump in RAM utilization for the large format time-lapse sequences (both playback and H.264).

                     

                    The biggest surprise for me was seeing the 9 x 5D MKII clips play perfectly simultaneously in the timeline (tiled) - this was ridiculously slow before my upgrade and would even cause crashes/hangs. I don't remember RAM utilization #s there but now am curious.

                    • 9. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                      Alex - DV411 Level 2

                      Ditto, thank you very much for posting your results, Luke.

                      At this modified size/quality setting - It works out to just under 1MB image vs. 6-7MB/image with higher quality settings, so at 24fps that's 24MB/s vs. 168 MB/s. Why 168 MB/s is too much for a disk system that reads up to 1 GB/s I'm not sure.

                      Possibly it's not the disk system but Premiere's ability to handle super-sized stills?  I am very curious, where the bottleneck is.

                      • 10. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                        Luke400 Level 1

                        So I followed up on the 9 clip sequence to see RAM utilization. It stayed under 6 GB, which I guess makes sense because the CPU would be the bottleneck here as it's decoding the h.264 MKII files. I should also add that the sequence does not playback smoothly, consistently. Some runs the playback will freeze up until I move the current marker in the timeline back then start again. I see the CPU getting pegged in these cases. Not sure why it works sometimes and not others. No other apps are open during these tests.

                         

                        RAM utilization goes up to 24 GB, however, when playing back the modified (increased compression) 2.5k JPEG time-lapses, so RjL I have to agree that 24 GB is probably a good number if you're bringing in time-lapses like this (or just generally for buffer). For other tests not involving these time-lapses, however, I didn't see RAM go much past 10 GB so 12-16 GB is probably OK.

                         

                        Message was edited by: Luke400

                        • 11. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                          ECBowen Most Valuable Participant

                          Cuda requires significant ram shadowing from system ram to GPU ram and then back after GPU processing. Because Parallel processing cannot share blocks of ram in GPU ram, this often means that as the data to the GPU grows, the ram exponentially grows and this allocation is outside of what Windows reports as ram usage since this is handled by the video driver. These memory blocks are dynamic and fluctuate constantly. The larger the frame resolutions to process by the GPU, the more ram blocks that will be created. This is why the Caching is changing so much with the MPE engine and also why one time on playback is smooth and the other time it stutters.

                           

                          Eric

                          ADK

                          • 12. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                            Luke400 Level 1

                            Thanks for the insight ECBowen.

                             

                            I have another update that is not good for me.

                             

                            1 - Very few of my video-based time-lapses will playback smoothly using native MKII files, none of them consistently. By "video-based time-lapses" I mean clips where I've taken video for 12+ minutes then sped it up to typically 10 seconds elapsed time in Premiere. It looks like I will have to continue to transcode to proxies for these files in order to work with them in the timeline. I use Neoscene for this, it's an unfortunate extra step.

                             

                            2 - Even the quality 60 JPEG time-lapses at 3840x2160 will blip every now and then ruining my flow when editing (noticing this as I try more of them). Rendering is super fast, but I have to re-render when I tweak pan/zoom. I will try slightly reducing the dimensions (giving up some pan/zoom latitude) to see if that helps.

                             

                            For video time-lapses (speed up native MKII files): Resource monitor does not show a disk bottle neck, and moving to my 1GB/s SSD raid doesn't help. CPU looks like the bottleneck here.

                             

                            So unfortunate news as this adds extra steps, many of them fairly time and disk consuming (in the case of proxies), but at least now I know the limitations. Would an SB 3860 proc help and/or a GTX 680? Not sure and no easy way to try.

                            • 13. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                              ECBowen Most Valuable Participant

                              Cineform Resamples the H264 files to 4 2 2 and is far easier for the player to process. That ultamately is a better workflow if you have effects, compositing, or color work on them. I would not see the Cineform conversion as a bad thing here.

                               

                              Eric

                              ADK

                              • 14. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                                Luke400 Level 1

                                Given the circumstances the Cineform tool is useful, but having to generate proxies in Cineform takes A LOT of time and ~2+ times the disk space of the original files. Plus I need to keep the originals. Overall a poor story vs. just using native files.

                                • 15. Re: Premiere Pro CS6: GTX 660 ti ($300)  vs. GTS 450 ($100) and other thoughts on upgrading HW
                                  hpmoon Level 1

                                  An extremely important (but rather obvious) clarification is that the only directly relevant test would be the GTS 450 installed versus the GTX 660 installed into the SAME faster system.