|
Testing results beta 1.00.14305 complete
I've done a lot of controlled "The Black Death" track replays, and since it's Monday in Russia - I thought I'd post my findings.
Here are graphs of 6 successive track playbacks attempting to isolate the cause of random stuttering and pausing on some track playbacks but not others. The following two examples iterate ProcessingAffinityMask and show the performance of CPU/GPU syncronized with other measurements and Fraps results. All my settings and system configuration is listed on these first two examples which display 22 minutes of data overall - each TBD track is 3 mins 40 secs in duration: Note: My HD5870 GPU is operating at maximum clock during loading from CoD track playback https://webspace.utexas.edu/joem/CLI...ad%20FINAL.jpg https://webspace.utexas.edu/joem/CLI...ad%20FINAL.jpg Note the playbacks in which the fps was 0 or lower than 10. Those are track playbacks with significant pausing and stuttering. Now here's a couple sets of measurements over two different tracks - at higher time resolution so you can see the relationship between GPU and CPU loading during a stutter/pause and a track with good performance - i.e. no stutter or pausing First the smooth track playback, note that GPU loading (purple) never drops lower than 50% during the track. https://webspace.utexas.edu/joem/CLI...erformance.jpg Now a playback which had significant pause (like 2 seconds). Note that during the stutter the CPU (yellow) drops to 0 first - then the GPU (purple) drops to 0. https://webspace.utexas.edu/joem/CLI...erformance.jpg No matter what Catalyst setting or CoD setting I changed - the stuttering and pausing continued randomly throughout testing. Many times, I thought one setting or the other had improved it - but follow on testing produced more stutter tracks. In summary - It's the CPU not the GPU that producing the stutter and pausing. As has been reported many times - the AA only works in two modes - OFF and 1X-2X-4X-8X all producing the same graphic output. Strangley, I found performance better at higher settings even though the screen output was the same. https://webspace.utexas.edu/joem/CLI...gsVsActual.jpg S! Gunny |
Nice work, saved the devs some work. Hopefully they read this and make some changes.
|
Nice work. I thought the big pauses were from the CPU being overloaded but it looks like otherway around - the game drops the ball and stops feeding it. :P
|
Thanks for the kind words. It took me days to figure this out, due to the randomness of runs. At one point I thought it was Tessellation, at another I thought it was ATITraytools interacting with the program, at one point I thought it was ProcessAffinity settings. All false leads
When I finally by the weekend, when I starting looking at the CPU/GPU interactions during stutters - I finally got it. It took all those plots and graphs to prove it however. Yes Phazon, that's exactly how it seems to me - the CPU gets starved for info and stops - then the GPU follows. In many runs you'll see the CPU start to come down in load - but it picks back up before the GPU is effected. I wanted to put this out here to stop folks from running out and buying video cards right now. By all means, wait until it all stabilizes then see what you need. Last thing: As you can see in the first series of graphs, my HD5870 is running at max clocks just fine. One of the interesting things that's occuring apparantely is the HD6000 series cards are not "clocking up". So HD5000's are operating one way - while HD6000's are operating in another. S! Gunny |
Great work Gunslinger, only downside is I think Russia is on 5 days off for May day holiday.
|
Quote:
http://forum.1cpublishing.eu/showpos...&postcount=289 It's the Ubi splash video that plays when the EU edition starts up (RU version doesn't have the Ubi splash so it's an EU problem only, afaik). With GPU-Z open at startup, you'll see normal clock speeds until that video plays - the vid causes the GPU clock to lock to the UVD video setting (usually 400mhz) and ATI powerplay won't detect the change, leaving us stuck at video clock speeds. Delete or rename the splash.wmv as described in the linked post and any cards suffering this issue should clock normally. :) |
I found that the game still trashing the HDD, even having free RAM to use, and that event cause shuttering.
If the game is not limited in the use of RAM and detects the available resources, and this objects are previously loaded in RAM, the HDD trashing shutters will disappears. Maybe is the search of this resources in the HDD what is causing this CPU hold back, or just is another source of shuttering different. |
Very good work!!
|
Quote:
It actually looks to me (if I had to be forced to guess) like a broken thread. i.e - one of the many processes that runs threaded, just had an error and stops. Then the process has to go and find something to "do". In short after several hundred runs this week, I saw no evidence of storage access issues. S! Gunny |
Quote:
Im talking about the first load in RAM, for example flying in "free fly mode" when you are approaching to a city the game shutters searching in the HDD some thing, a house or something, when it is already in RAM there no shutter flying the same zone. But there free RAM, why is not this loaded already ? :confused: About the strange issue that there better performance with AA in x8 than in x4. That is a ATI issue, they have better performance in x8 than x4, even if it is not working properly, dont know why :confused: |
Quote:
Also, there's a part of TBD track (later portion of track) in which the camera zooms into a RAF Flak Gunner. In almost every (if not all) of the intial passes after restarting CoD - the animation would be missing ammunition load in the AA gun. All other passes - the ammo showed up and as the gun fired the ammo load decreased. I hate that Flak Gunner by the way. He has a bug. His vertical position is too low - so that the AA Guns seat - passes through is neck - instead of his butt. I'll be glad when they fix that and I won't have to look at it anymore. S! Gunny |
Great job Gunny, thank you!
|
Interesting research, and amazing scientist attitude. Good.
|
thx for the hard work gunny
|
Thanks Guys for reading through all that and for the kind words. I'm glad you find it useful. At one point last week I was saying to myself "stop this benchmarking instanity and go practice dropping bombs and shooting", because I wasn't getting anywhere.
I thought on Sunday morning, sheesh - relook at the results and see if there's anything in this that would help anyone and I noticed then the GPU/CPU timing in the ATITrayTools graphs. As an aside - here's a dump of one of my high level logs where I've complied results and differences in settings from the initial baseline - these runs reflect about 1-2 testing days: Look at the variabilty throughout the multiple pass runs. Always there. Once time it's not but repeat the same settings and the hard stutter will find a way to come back. You can only see this type of behavior in multiple iterations - 3 of 5 track runs - etc.... this is just Sat and Sun's runs :) AAOFF-Ap-Morphon 1.00.14101 Frames, Time (ms), Min, Max, Avg 8884, 210143, 10, 82, 42.276 1.00.14214 Frames, Time (ms), Min, Max, Avg 6281, 213048, 0, 83, 29.482 AAOFF-AP-MorhOFF 1.00.14101 Frames, Time (ms), Min, Max, Avg 9188, 213827, 6, 110, 42.969 1.00.14214 Frames, Time (ms), Min, Max, Avg 8688, 213767, 0, 87, 40.642 4XAA in AP - Morphon 1.00.14101 Frames, Time (ms), Min, Max, Avg 5750, 211313, 0, 108, 27.211 1.00.14214 Frames, Time (ms), Min, Max, Avg 5925, 213186, 0, 69, 27.793 4XAA in AP - MorphOFF 1.00.14101 Frames, Time (ms), Min, Max, Avg 5766, 217886, 0, 108, 26.463 1.00.14214 Frames, Time (ms), Min, Max, Avg 8465, 214107, 0, 88, 39.536 1.00.14305 GrassOFF 4XAA AP/ED - AF Appcontrolled - Grass OFF|Tesselation 64X|Texture Filter High Qual - NO Surface Format Optimization - Adaptive MS Frames, Time (ms), Min, Max, Avg 8893, 212102, 15, 98, 41.928 Same as above - now add 8X AF Frames, Time (ms), Min, Max, Avg 9010, 215613, 15, 97, 41.788 AF to AP - 0XAA in APcontrol - all other settings the same Frames, Time (ms), Min, Max, Avg 8659, 213278, 0, 89, 40.600 Frames, Time (ms), Min, Max, Avg 8789, 213669, 2, 89, 41.134 2XAA in AP Consecutive runs taken after previous runs with many screensnaps Frames, Time (ms), Min, Max, Avg 7956, 210537, 0, 91, 37.789 Frames, Time (ms), Min, Max, Avg 7780, 212982, 0, 88, 36.529 Post reboot Frames, Time (ms), Min, Max, Avg 8672, 211015, 0, 95, 41.097 8XAA in AP Frames, Time (ms), Min, Max, Avg 9136, 213765, 16, 98, 42.739 8XAA in AP - Tesselation = 64 Frames, Time (ms), Min, Max, Avg 8953, 212338, 18, 99, 42.164 8XAA in AP - Tessellation = OFF Frames, Time (ms), Min, Max, Avg 8478, 212003, 0, 94, 39.990 Frames, Time (ms), Min, Max, Avg 8731, 212338, 0, 94, 41.118 Frames, Time (ms), Min, Max, Avg 8732, 213975, 0, 97, 40.809 8XAA in AP - Tessellation = 64X Frames, Time (ms), Min, Max, Avg 8953, 212338, 18, 99, 42.164 8XAA in AP - Tessellation = 64X - ATITrayTools Running Frames, Time (ms), Min, Max, Avg 8787, 211340, 0, 96, 41.578 2 big pauses 8830, 211417, 7, 95, 41.766 2 stutters 8569, 211280, 0, 94, 40.558 pauses 8043, 211964, 0, 92, 37.945 8 stutters 8578, 211848, 0, 91, 40.491 pauses 8XAA in AP - Tessellation = 64X - ATITrayTools Shutdown Frames, Time (ms), Min, Max, Avg 9114, 212280, 17, 98, 42.934 Smooth 8994, 212468, 16, 96, 42.331 Smooth 9093, 212634, 17, 96, 42.764 Smooth 9000, 212428, 15, 97, 42.367 Smooth 9007, 212326, 17, 97, 42.421 Smooth 8xAA in AP - Tessellation = OFF - ATITrayTools Shutdown 9012, 212089, 16, 97, 42.492 Smooth 8905, 212175, 17, 93, 41.970 Smooth 8784, 208758, 17, 95, 42.077 boringly smooth 8822, 212061, 16, 93, 41.601 Smooth 8XAA in AP - Tessellation = OFF - ATITrayTool Running 8754, 212337, 0, 96, 41.227 large pauses 8892, 212400, 18, 97, 41.864 smooth! 8839, 211994, 14, 93, 41.695 smooth 8414, 212259, 0, 93, 39.640 8614, 212233, 0, 95, 40.587 8605, 212233, 0, 92, 40.545 8562, 212274, 0, 89, 40.335 8688, 212419, 9, 91, 40.900 some jerkyness 8768, 212443, 16, 95, 41.272 smooth 8571, 212186, 0, 95, 40.394 OK now - everything same. ATITrayTool Shutdown. HP Button Manager ON 8XAA in AP Catalyst 11-4 Frames, Time (ms), Min, Max, Avg 8986, 210598, 16, 97, 42.669 couple micro-stutters 8767, 211908, 0, 97, 41.372 1 micro-pause + 1 Huge Pause 8579, 213074, 0, 99, 40.263 2 large stutters and a pause 8297, 212141, 0, 94, 39.111 Pause X 3 8725, 212464, 0, 99, 41.066 Pause X 3 + 1 Huge Pause 1 run with HP Button Manager off after reboot - first run had to be aborted due to severe pausing at start of 2nd (monitored) run of track. No change. 9042, 213308, 0, 98, 42.389 8XAA in AP - ATITrayTool Running - Process Affinity Mask = 15 in /Softclub path conf.ini Frames, Time (ms), Min, Max, Avg 9042, 213308, 0, 98, 42.389 Initial Load 9351, 212584, 0, 101, 43.987 8439, 212245, 0, 90, 39.761 8712, 212537, 0, 96, 40.991 8704, 212117, 2, 95, 41.034 8XAA in AP - ATITrayTool Running - Process Affinity Mast = 14 8601, 212202, 17, 91, 40.532 8550, 209688, 0, 91, 40.775 8682, 212022, 16, 92, 40.949 8575, 211675, 0, 92, 40.510 8XAA in AP - ATITrayTool Running - Process Affinity Mask OFF Frames, Time (ms), Min, Max, Avg 9333, 211847, 18, 106, 44.055 - Initial Load Run 8831, 212286, 15, 98, 41.600 8819, 212457, 16, 98, 41.510 8710, 212362, 13, 95, 41.015 8713, 212642, 15, 95, 40.975 Same exact configuration as above - reboot to confirm next runs 9538, 212237, 6, 106, 44.940 9119, 212725, 15, 101, 42.868 9001, 212589, 18, 98, 42.340 8901, 212659, 0, 95, 41.856 8902, 212663, 17, 99, 41.860 8968, 212485, 3, 100, 42.205 Process Affinity Mask set to = 15 Frames, Time (ms), Min, Max, Avg 9392, 212447, 0, 106, 44.209 9009, 211982, 0, 99, 42.499 9047, 212045, 15, 96, 42.665 8920, 212004, 18, 93, 42.075 8999, 213199, 18, 95, 42.209 8807, 212720, 0, 97, 41.402 S! Gunny |
Great job and a constructive attitude towards problems. This doesn't only help the developers if they don't already know the cause of what they are trying to fix, it also helps the users apply possible work-arounds and decide if they should upgrade or refrain from throwing money and hardware at the problem if it's something fixable.
Much more useful than the usual "it's broken, fix it" one-liners we so often get ;) |
Quote:
Many thanks for this great work! ~S~ |
Good work. I don't have the game, but appreciate this anyway.
|
Very nice work Gunslinger!
|
Just to say hello to an old friend from RoF core beta test ;)
Excellent work Gunny! |
Do you use an SSD in your system? Know you very much doubt a harddisk is to blame for the stutters, but as far as I know textures get loaded an unloaded into the cache as they are needed. So unless you have a very large pagefile and lots of RAM AND CLOD is using it all instead of filling to a max of say 2GB before reusing it it seems to me as the most obvious reason for making the CPU wait.
Just asking because I built a new pc recently with CLOD installed on a Crucial C300 SSD and am experiencing no stutters whatsoever and other people with SSD's seem to experience little to no stutters aswell. |
If you dont have a fast HDD but do have 6+ Gig of Ram you can just disable your page file altogether. It´s what I did and it works fine.
|
Quote:
All the best, Gunny |
Quote:
Thanks Gunny |
Quote:
Important to say here, many tracks had no or very little stutter. When I play the game - versus test a high load performance track - I don't have a lot of stutter complaints. If you look back at my results - many times the lowest fps in tracks is 14-to-18 with no or minimal stutter. Repeated playbacks and some initial playback - always eventually show it in a random manner. By random I mean - at different spots on the track. Sometimes the stutter occurs when the attacking 110's are viewed from the perspective of the Hurricanes waiting to take off - sometimes it occurs a low load points on the track - it's random. The pauses can also be very short and sometimes as long as 2-3 seconds. This is a problem with execution, which I'm sure will be nailed down soon. When you guys are saying you have found solutions that have no stutter - Do you mean normal execution of the Simulator - or do you mean repeated playback of "The Black Death" track? Two different conditions. Try 7 sequential playbacks of "The Black Death" track without restarting CoD. If you get through that with no Minimum FPS less than 10 fps and I'm pretty sure you have found a solution. I had one or two examples in a week of testing - in which I had 4 or 5 consecutive runs with minimum FPS of 14fps or above. However, the rub is that after both those episodes, when I tried to re-run the next 5 runs with the same settings after restarting CoD - the stutters returned. Thanks Gunny |
Quote:
|
Quote:
Warning here though, only do this if you have plenty RAM (in your case this shouldnt be a problem at all). This is of course totally reversible should you not like it after all. Right here goes: Go to Control Panel -> System & Security -> System -> Advanced Settings (left side of panel, run as admin) -> Advanced panel -> Go to top tab there and pick definitions -> You will see a tab with three options (Visual Effects, Advanced and one other) -> Pick Advanced -> Virtual Memory -> Change -> Untick the Automatically set the pagefile folder thingy -> Select the No PageFile thingy -> Profit. Sorry for the somewhat vague walkthrough but it should get you there. Regards. |
S!
I run both SSD and without pagefile. No problems so far. Great testing Gunny. I had no probs with the 3D clocks after I overclocked my card a whopping 5MHz :D Let's see what the patch will do.. |
It's my opinion that the stutters don't come from the HDD (maybe there are exceptions). I've tested the game on a HDD (7200) and on 3 very fast SSD's in a RAID 0 array (which is probably bottlenecked by the controller or bus). It's my experience that it makes totally no difference in stuttering. Thereby I have to say that my stutters are mostly mild.
I agree with TUSA/TX-Gunslinger that it has probably something to do with the code of the game. On the moments that you see a short heavy drop in FPS you see a simultaneous (more or less) drop in CPU utilization. Normally CPU utilization should increase when a game has lots to process. |
Hi all,
Thanks for testing! BTW, what tool did you use to measure the GPU utilization? Leo "Apollo11" |
Quote:
Any graphs or screens that show this effect in action? Obviously you can't show us the stutters, but a CPU monitor shot showing these dropouts could go a long way in helping the devs (or us!) solve the stutter problem? |
ATITrayTools - found under the Hardware Tab. You can set it to monitor all sorts of things from your AMD/ATI GPU ......... I love the flexiblity of the graphs. As you can see by my examples - you can control X axis (time) and Y axis (measurement value) independently - which allows you see all sorts of interactions.
Anyway, here's a link to guru3d http://www.guru3d.com/article/ati-tray-tools-/ Oh yes, another benie.... it's free :) S! Gunny |
Hi all,
Quote:
Leo "Apollo11" |
Outstanding work Gunny. :)
This thread makes for most interesting reading. |
Quote:
For me it's worth the extra $150 for the faster system startup and game load times. And I'm not even part of the 'Now' generation. Lifes to short to be waiting around for your PC to start. Cheers! |
If the cpu stops to wait on something(thus causing stutter) would it be better to under clock the cpu?
|
My CPU fan goes mental when im playing this game with the newest beta patch...as soon as i pause game or exit you can hear it start to spin down..
|
Quote:
|
| All times are GMT. The time now is 08:18 AM. |
Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.
Copyright © 2007 Fulqrum Publishing. All rights reserved.