Page 3 of 10 FirstFirst 123456 ... LastLast
Results 51 to 75 of 236

Thread: HD5970 Microstuttering tests

  1. #51
    Xtreme Mentor
    Join Date
    May 2008
    Location
    cleveland ohio
    Posts
    2,879
    Quote Originally Posted by Particle View Post
    I'm not sure I understand your question. If you're asking if there are more modes than just AFR and supertiling, yes, there are. Supertiling is pretty much the end-game as far as getting rid of visual problems though, so I recommend it.
    yes exactly, but you need more then 2 card to get those options to show up.
    HAVE NO FEAR!
    "AMD fallen angel"
    Quote Originally Posted by Gamekiller View Post
    You didn't get the memo? 1 hour 'Fugger time' is equal to 12 hours of regular time.

  2. #52
    Xtreme Enthusiast
    Join Date
    Aug 2008
    Posts
    577
    I'm still baffled that 95% of people claim "microstutter" when, it is just stutter. "Microstutter" was an effect where variations in frame rendering time would become even more fubar when using different refresh rates than what the monitor would support...or something to that affect.

    Sutter != microstutter

    I should also add that you should be testing with either Core i7 or Phenom II, as even noticed by Anandtech gameplay actually seemed smoother in many games, even when the FPS didn't show it.
    --Intel i5 3570k 4.4ghz (stock volts) - Corsair H100 - 6970 UL XFX 2GB - - Asrock Z77 Professional - 16GB Gskill 1866mhz - 2x90GB Agility 3 - WD640GB - 2xWD320GB - 2TB Samsung Spinpoint F4 - Audigy-- --NZXT Phantom - Samsung SATA DVD--(old systems Intel E8400 Wolfdale/Asus P45, AMD965BEC3 790X, Antec 180, Sapphire 4870 X2 (dead twice))

  3. #53
    Xtreme Addict
    Join Date
    Mar 2008
    Location
    川崎市
    Posts
    2,076
    Quote Originally Posted by STaRGaZeR View Post
    The human eye can't see more than 60FPS anyway
    Regardless if it can or not, think about the max number the majority of TFTs on the market today actually displays (and you can definitly not see more than the monitor displays).

    Anyway, there are three problems that lead to the user experiencing stuttering, crappy min fps, input lag and micro stuttering, only one of them is exclusive to multi gpu setups and can be reduced by not using AFR, so why is everyone bashing on multi gpu setups like they would be the spawn of evil?
    Last edited by naokaji; 01-15-2010 at 01:48 AM.

  4. #54
    Xtreme Mentor
    Join Date
    Nov 2006
    Location
    Spain, EU
    Posts
    2,949
    Quote Originally Posted by naokaji View Post
    Regardless if it can or not, think about the max number the majority of TFTs on the market today actually displays (and you can definitly not see more than the monitor displays).
    I know that, that's why some of us have one of those 120Hz monitors and are enjoying them like if it were no tomorrow. Too bad they're only TN at the moment.
    Friends shouldn't let friends use Windows 7 until Microsoft fixes Windows Explorer (link)


    Quote Originally Posted by PerryR, on John Fruehe (JF-AMD) View Post
    Pretty much. Plus, he's here voluntarily.

  5. #55
    Xtreme Member
    Join Date
    Dec 2008
    Location
    Washington DC
    Posts
    302
    Quote Originally Posted by anandtech
    However even as simple as AFR is, it isn’t foolproof and it isn’t flawless. Making it work at its peak level of performance requires some understanding of the game being run, which is why for even such a “dumb” method we still have game profiles. Furthermore it comes with a few inherent drawbacks

    1. Each GPU needs to get a frame done in the same amount of time as the other GPUs.
    2. Because of the timing requirement, the GPUs can’t differ in processing capabilities. AFR works best when they are perfectly alike.
    3. Dealing with games where the next frame is dependent on the previous one is hard.
    4. Even with matching GPUs, if your driver gets the timing wrong, it can render frames at an uneven pace. Frames need to be spaced apart equally – when this fails to happen you get microstuttering.
    5. AFR has multiple GPUs working on different frames, not the same frame. This means that frame throughput increases, but not the latency for any individual frame. So if a single card gets 30fps and takes 16.6ms to render a frame, a pair of cards in AFR get 60fps but still take 16.6ms to render a frame.
    http://www.anandtech.com/video/showdoc.aspx?i=3713&p=2
    Good explanation of what happens. And like many things, multi-gpu setups are user preference.

    I never was happy with my 4870 xfire, it felt sluggish to me.
    Last edited by Sgt.McRuff; 01-14-2010 at 11:28 AM.
    Motherboard: GigaByte P67UD4 f6 | CPU: Intel 2500k 4.5ghz 1.26v | Memory: GSkill 2x4gb @ 1600mhz 1.34v | PSU: SeaSonic X650 Gold 650W | Video: AMD 6970 Koolance water block 880c/1450mem 1.035v | HDD: WD 640gb cavier black: VelociRaptor 300gb: Intel x-25 g2 80gb | Sound: Asus xonar D1 | OS: W7 64bit

  6. #56
    Xtreme Addict
    Join Date
    Aug 2007
    Location
    Istantinople
    Posts
    1,574
    The people who don't notice microstuttering are playing games in which the minimum momentary FPS doesn't go below 50 or 60 something. When your momentary FPS is 50-100-50-100-50-100 you don't see a 75 FPS level fluidity, you see 50 FPS level fluidity. But when your FPS's are so high that you don't see microstuttering, then your second GPU doesn't add anything, does it? All you have done is to pay huge amounts of extra money just to see that 75 FPS indicator whereas you are getting a mere 50 FPS.

    I am not sure that this is a "feeling" issue that varies from person to person. Sure, min acceptable / perceived FPS varies from person to person but that's it. For everyone, the perceived FPS has to be the low momentary FPS. Just think this way - every even frame is rendered at 10ms (100 FPS) and every odd frame at 1000ms (1 FPS). So the frame times go 10-1010-1020-2020 etc. Would you perceive 50 FPS (the average) here, or 1 FPS?
    Has anyone really been far even as decided to use even go want to do look more like?
    INTEL Core i7 920 // ASUS P6T Deluxe V2 // OCZ 3G1600 6GB // POWERCOLOR HD5970 // Cooler Master HAF 932 // Thermalright Ultra 120 Extreme // SAMSUNG T260 26"

  7. #57
    Xtreme Enthusiast TJ TRICHEESE's Avatar
    Join Date
    May 2009
    Posts
    784
    i have to disagree with the general direction of this argument as i have had much better performance with my gtx285 tri-sli, less MS than with one 9800gx2 or two gtx 280's, im not saying that our current method of multi-gpu rendering is perfect, im just saying that not all setups with multiple cards are useless, some setups actually work as average fps (including MS) is increased with more cards

  8. #58
    Xtremely Kool
    Join Date
    Jul 2006
    Location
    UK
    Posts
    1,875
    Quote Originally Posted by TRICHEESE View Post
    i have to disagree with the general direction of this argument as i have had much better performance with my gtx285 tri-sli, less MS than with one 9800gx2 or two gtx 280's, im not saying that our current method of multi-gpu rendering is perfect, im just saying that not all setups with multiple cards are useless, some setups actually work as average fps (including MS) is increased with more cards
    Indeed the same multi gfx cards in another setup will most likely give different MS results.

  9. #59
    Xtreme Enthusiast Kai Robinson's Avatar
    Join Date
    Oct 2007
    Location
    East Sussex
    Posts
    831
    I've got two HD4850 cards in crossfire. In 99.9% of games, i don't notice any 'microstuttering' with my Q9650 @ 4Ghz (P5Q Deluxe Board), except Crysis - but then, ATi + Crysis + Crossfire = Fail combination, thanks to Cryteks sloppy coding.

    However, when i had an E7200 @ 4Ghz, it was more noticeable in Crysis - personally, i think it's probably more a case of the CPU not having enough grunt to farm out data fast enough.

    Fallout 3 @ High details, 1680x1050 with one 4850 is damn near unplayable. Only by adding a 2nd card, do i see more than 30fps (max of about 70fps). Performance dips in texture heavy outdoors scenes in either case, but no stuttering occurs.

    Maybe i'm just old and my eyes are slow...?

    Main Rig

    Intel Core i7-2600K (SLB8W, E0 Stepping) @ 4.6Ghz (4.6x100), Corsair H80i AIO Cooler
    MSI Z77A GD-65 Gaming (MS-7551), v25 BIOS
    Kingston HyperX 16GB (2x8GB) PC3-19200 Kit (HX24C11BRK2/16-OC) @ 1.5v, 11-13-13-30 Timings (1:8 Ratio)
    8GB MSI Radeon R9 390X (1080 Mhz Core, 6000 Mhz Memory)
    NZXT H440 Case with NZXT Hue+ Installed
    24" Dell U2412HM (1920x1200, e-IPS panel)
    1 x 500GB Samsung 850 EVO (Boot & Install)
    1 x 2Tb Hitachi 7K2000 in External Enclosure (Scratch Disk)


    Entertainment Setup

    Samsung Series 6 37" 1080p TV
    Gigabyte GA-J1800N-D2H based media PC, Mini ITX Case, Blu-Ray Drive
    Netgear ReadyNAS104 w/4x2TB Toshiba DTACA200's for 5.8TB Volume size

    I refuse to participate in any debate with creationists because doing so would give them the "oxygen of respectability" that they want.
    Creationists don't mind being beaten in an argument. What matters to them is that I give them recognition by bothering to argue with them in public.

  10. #60
    Xtreme Guru
    Join Date
    May 2007
    Location
    Ace Deuce, Michigan
    Posts
    3,955
    Honestly why don't we just let those who don't experience ms enjoy their gift? Think about it, that's one of the few times it actually improves your experience to not be able to detect the effects of your hardware and can be seen as a blessing considering how much money they're spending.

    I just don't see any point to arguing over something as mundane as that; if anything keep it strictly between people who do experience ms as those who don't can't comment from experience anyways.
    Quote Originally Posted by Hans de Vries View Post

    JF-AMD posting: IPC increases!!!!!!! How many times did I tell you!!!

    terrace215 post: IPC decreases, The more I post the more it decreases.
    terrace215 post: IPC decreases, The more I post the more it decreases.
    terrace215 post: IPC decreases, The more I post the more it decreases.
    .....}
    until (interrupt by Movieman)


    Regards, Hans

  11. #61
    I am Xtreme
    Join Date
    Dec 2007
    Posts
    7,750
    Quote Originally Posted by AliG View Post
    Honestly why don't we just let those who don't experience ms enjoy their gift? Think about it, that's one of the few times it actually improves your experience to not be able to detect the effects of your hardware and can be seen as a blessing considering how much money they're spending.

    I just don't see any point to arguing over something as mundane as that; if anything keep it strictly between people who do experience ms as those who don't can't comment from experience anyways.
    simple, those who dont notice it dont have to worry about their 600$ card feeling like crap. the rest of the world probably wants to see what causes it and if there are ways around it (platform, resolution, AA, etc)

  12. #62
    Xtremeish
    Join Date
    Nov 2005
    Location
    Norway
    Posts
    1,190
    I also thought this was a thing of the past, however I have noticed exactly this in Fallout 3 and it bugs me.
    Other games I can't say I remember, but that's probably because my mind is completely elsewhere.
    One thing is sure though, until they fix this I am not buying another dual GPU card.
    Aber ja, naturlich Hans nass ist, er steht unter einem Wasserfall - James May
    Hardware: Gigabyte GA-Z87M-D3H, Intel i5 4670k @ 4GHz, Crucial DDR3 BallistiX, Asus GTX 770 DirectCU II, Corsair HX 650W, Samsung 830 256GB, Silverstone Precision -|- Cooling: Noctua NH-C12P SE14

  13. #63
    Xtreme Member
    Join Date
    Dec 2005
    Posts
    427
    In my case I think SLI works without MS in case I enable Vertical sync and triple buffering.

  14. #64
    Registered User
    Join Date
    Aug 2009
    Posts
    20
    I always play with VSYNC enabled, and frame rates locked to 60Hz. If you worked with interlaced movies, bad field order, converting material from NTSC to PAL, then you know what stuttering is. Monitors cant delay pictures 1 or 2 msec, they work with 20 ms or 16,666 msec timebase. The only way to see 1-2 msec stutter is, if the game engine is crap, and cant properly track its internal timer to render the frames properly, or uses RDTSC instructions for timing from 2-3-4 different cores. So, i think the SLI/CF technology is satisfactory so far, only the GAMES are very poorly ported/written.

  15. #65
    Xtreme Member
    Join Date
    Jul 2006
    Location
    Ontario, Canada
    Posts
    251
    Microstutter is solved simply by using vsync. I don't understand why anybody in 2010 is playing games without vsync, putting up with horrible image tearing and uneven frame times.

    The number of FPS on your screen means NOTHING, it's the time between frames that matters, period.

    It's like me saying I give you 7 chickens to eat in a week, would you rather have a fresh hot chicken everyday, or 3 chickens monday and 4 chickens sunday?

    Vsync is the rationing agent, one frame every 16.6 ms, it should be mandatory, it should be enabled by default in the driver and never allowed to be disabled.

    What's sad is that we're in 2010, and we still don't have a system built-in the drivers to remove input lag that comes with vsync. The good old 59 fps cap works in most cases though.

    Intel Core i7 3770K | Asus Maximus V Gene | 16GB DDR3-1600 | Asus GTX 670 directCU TOP | Intel 320 120GB | WD Caviar Black 1TB | Enermax Revolution 1050W | Silverstone TJ08-E | Dell UltraSharp U2711

  16. #66
    Xtreme Addict
    Join Date
    Jan 2008
    Location
    Puerto Rico
    Posts
    1,374
    Quote Originally Posted by JAG87 View Post
    Microstutter is solved simply by using vsync. I don't understand why anybody in 2010 is playing games without vsync, putting up with horrible image tearing and uneven frame times.

    The number of FPS on your screen means NOTHING, it's the time between frames that matters, period.

    It's like me saying I give you 7 chickens to eat in a week, would you rather have a fresh hot chicken everyday, or 3 chickens monday and 4 chickens sunday?

    Vsync is the rationing agent, one frame every 16.6 ms, it should be mandatory, it should be enabled by default in the driver and never allowed to be disabled.

    What's sad is that we're in 2010, and we still don't have a system built-in the drivers to remove input lag that comes with vsync. The good old 59 fps cap works in most cases though.
    and now more than ever you can vsync + triple buffering = win with 120hz displays!
    ░█▀▀ ░█▀█ ░█ ░█▀▀ ░░█▀▀ ░█▀█ ░█ ░█ ░░░
    ░█▀▀ ░█▀▀ ░█ ░█ ░░░░█▀▀ ░█▀█ ░█ ░█ ░░░
    ░▀▀▀ ░▀ ░░░▀ ░▀▀▀ ░░▀ ░░░▀░▀ ░▀ ░▀▀▀ ░

  17. #67
    c[_]
    Join Date
    Nov 2002
    Location
    Alberta, Canada
    Posts
    18,728
    Quote Originally Posted by Lokinhow View Post
    oh maan...

    multi GPU is around for yeeears, but it doesn't work properly yet


    ok, AFR is crap, so what would be the "right way" to do multi GPU?
    SLI (Scan Line Interleve - 3DFX style), SFR (Split-Frame Rendering, true split or work-load balanced) and Micro/Macro-Tiling (remember that tweak from R100 days?).


    edit - Vysnc sucsk btw, adds latency to image display.

    All along the watchtower the watchmen watch the eternal return.

  18. #68
    Xtreme Enthusiast
    Join Date
    Mar 2009
    Posts
    605
    Beer, the MS cure. Seems to work every time.
    Email/web browsing PC
    MM Ascension case, 24/7 settings
    Classified, 980x @ 4,525Mhz, HT on 1.39V
    (3) ATI 5870's@ 1062Mhz
    G skill perfect storm RAM s 6gb, 2,176 Mhz CAS 8
    (4) intel x25m RAID 0
    CPU loop, EK supreme HF, EK mobo block, 500MM bitspower res, 2 DDC 3.25's on XSPC top, 2 XSPC RX 360's, 1 XSPC RX 480, GPU loop, EK water blocks, with nickel backs, 2 swiftech mcr 320's, 500MM bitspower res, ddc 3.25, XSPC top all with 1850 RPM gentle typhoons
    plus many other rigs.

  19. #69
    Xtreme Member
    Join Date
    Feb 2007
    Posts
    337
    Quote Originally Posted by STEvil View Post
    SLI (Scan Line Interleve - 3DFX style), SFR (Split-Frame Rendering, true split or work-load balanced) and Micro/Macro-Tiling (remember that tweak from R100 days?).


    edit - Vysnc sucsk btw, adds latency to image display.
    This so reminds me of the debate I was in when nvidia's SLi was announced...I didn't really like what I saw. I saw a half-assed approach to going back to nostalgic multi-card performance...only with less than nostalgic performance gains.

    People told me it was because hardware and bandwidth limitations no longer allowed the old Scan-Line-Interleave way to work....

    I'm not saying modern SLi doesn't have it's value. But it sure could be a lot better if it was like the original... If only they could develop an interconnect to make this happen.
    Last edited by Rock&Roll; 01-14-2010 at 09:03 PM.
    System: Core I7 920 @ 4200MHz 1.45vCORE 1.35VTT 1.2vIOH // EVGA x58 Classified E760 // 6GB Dominator GT 1866 @ 1688 6-7-6-18 1T 1.65V // Intel X25 80GB // PCP&C 750W Silencer
    Cooling: Heatkiller 3.0 LT CPU block // 655 Pump // GTX360 Radiator
    Sound: X-FI Titanium HD --> Marantz 2265 --> JBL 4311WXA's
    Display: GTX480 // Sony GDM-FW900

  20. #70
    Xtreme Member
    Join Date
    Jan 2009
    Location
    Huyamba
    Posts
    316
    Quote Originally Posted by Rock&Roll View Post
    This so reminds me of the debate I was in when nvidia's SLi was announced...I didn't really like what I saw. I saw a half-assed approach to going back to nostalgic multi-card performance...only with less than nostalgic performance gains.

    People told me it was because hardware and bandwidth limitations no longer allowed the old Scan-Line-Interleave way to work....

    I'm not saying modern SLi doesn't have it's value. But it sure could be a lot better if it was like the original... If only they could develop an interconnect to make this happen.
    +1
    Everything that nvidia has "invented" - was actually the 3Dfx heritage(well almost everything). Now it has to do something on its own - and oops - it turns out to be so hard to do. Delays, delays - especially when you are not alone on the market and can't sell the same thing everytime but instead need to make smth actually new...
    i7 950@4.05Ghz HeatKiller 3.0
    EVGA E762 EK WB | 12Gb OCZ3X1600LV6GK
    Razer Tarantula |Razer Imperator | SB X-Fi PCIe
    480GTX Tri SLi EK WBs | HAF X | Corsair AX1200
    ____________________________________________
    Loop1: Double_MCP655(EK Dual Top) - MoRa3Pro_4x180 - HK3.0 - EKFB_E762
    Loop2: Koolance_MCP655(EK Top) - HWLabsSR1_360 - EK_FC480GTX(3x)

  21. #71
    Xtreme X.I.P.
    Join Date
    Nov 2002
    Location
    Shipai
    Posts
    31,147
    Quote Originally Posted by Shadov View Post
    I disagree saaya, some titles are more prone to this effect then others.

    That way you CAN NOT make it a general rule for all titles and just support it by a fraps graph.
    true... it seems it depends on the games and the framerate (the lower the worse it gets)

    Quote Originally Posted by Lokinhow View Post
    ok, AFR is crap, so what would be the "right way" to do multi GPU?
    not really... if you do afr in a smart way it should work flawlessly... you just need to align the 2 or more gpus properly so they steadily fire fps.

    Quote Originally Posted by Eastcoasthandle View Post
    Folks, just don't read the OP but click on the link and read all the graphs. There you will find a trend where fluctuates are more pronounced at 1920 then at 1680. Thus it's easy to see that there is some sort of frame rate induced problem more so then anything else.
    well yeah, but who will play at 1680 with a 5970?

    Quote Originally Posted by Sentential View Post
    The worst games for microstuttering in my personal experience have been MMOs and games that have large open worlds that have to access heavy amounts of CPU/Mem/HDD loads.
    well are you sure its microstuttering and not the game running out of frame buffer and having to cache data from the system mem or hdd? that causes pretty bad stuttering

    its possible that this then triggers microstuttering as well though, as all gpus run out of data and then start rendering again at almost the same time, causing very uneven frame times.

  22. #72
    Xtreme X.I.P.
    Join Date
    Nov 2002
    Location
    Shipai
    Posts
    31,147
    Quote Originally Posted by KingOfsorroW View Post
    2all I don't know, I'm not blind and definitely saw no microstuttering - everything was smooth like baby's a$$ with that 3 sli setup.
    Macro stuttering - constant big fluctuations (the usual stuttering in another words)
    microstuttering - small and frequent fluctuations (may happend randomly).
    There was a lot of talks about microstuttering on the internet at the time of gtx295 -285 launch - the microstuttering is a myth. Because it is almost impossible to catch it for a human eye. People tend to confuse it with some other troubles like latency or the margin between the min and max fps for example...
    But one thing for sure - one fast card is better than 2 or 3 cards alright.
    right, so if you cant see it, it doesnt exist... :P

    Quote Originally Posted by JAG87 View Post
    Microstutter is solved simply by using vsync.
    err no? vsync causes big problems with some cases... read the anandtech article about it, i think it was called gaming latency explored or something like that...

    Quote Originally Posted by justin.kerr View Post
    Beer, the MS cure. Seems to work every time.
    heheh, true

  23. #73
    Xtreme Member
    Join Date
    Jan 2009
    Location
    Huyamba
    Posts
    316
    Quote Originally Posted by saaya View Post
    right, so if you cant see it, it doesnt exist... :P
    With vsync on - no, it doesn't.
    i7 950@4.05Ghz HeatKiller 3.0
    EVGA E762 EK WB | 12Gb OCZ3X1600LV6GK
    Razer Tarantula |Razer Imperator | SB X-Fi PCIe
    480GTX Tri SLi EK WBs | HAF X | Corsair AX1200
    ____________________________________________
    Loop1: Double_MCP655(EK Dual Top) - MoRa3Pro_4x180 - HK3.0 - EKFB_E762
    Loop2: Koolance_MCP655(EK Top) - HWLabsSR1_360 - EK_FC480GTX(3x)

  24. #74
    Xtreme Addict
    Join Date
    Jul 2007
    Location
    Alberta, Canada
    Posts
    1,264
    Quote Originally Posted by Beefy22 View Post
    Is that MS or another phenomena altogether?
    Uneven frametimes so yes. I'm much happier on a single gpu.

    As far as the whole Vsync debate, personally I can put up with it in single player games but in 99% of online first person shooters, vsync = epic fail. Enough said. If you don't know what I am speaking of then clearly you don't play online heheh...
    Feedanator 7.0
    CASE:R5|PSU:850G2|CPU:i7 6850K|MB:x99 Ultra|RAM:8x4 2666|GPU:980TI|SSD:BPX256/Evo500|SOUND:2i4/HS8
    LCD:XB271HU|OS:Win10|INPUT:G900/K70 |HS/F:H115i

  25. #75
    c[_]
    Join Date
    Nov 2002
    Location
    Alberta, Canada
    Posts
    18,728
    This isnt true for all situations, but i've noticed a lot engines that have micro-stutter issues seem to stream a lot of data due to lack of VRAM or they just need to stream data due to sheer amount to show the entire game world.

    Add in some fast paces and you can really see the macro-stutter effect (pauses)...

    All along the watchtower the watchmen watch the eternal return.

Page 3 of 10 FirstFirst 123456 ... LastLast

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •