What like FPS? That's been tested for years and it certainly isn't noticeable past a certain point, unless there is stuttering . . . .
Seems to me like TR has developed a useful tool here.
Last edited:
What like FPS? That's been tested for years and it certainly isn't noticeable past a certain point, unless there is stuttering . . . .
Seems to me like TR has developed a useful tool here.
njdevilsfan87
Platinum Member
- Apr 19, 2007
- 2,341
- 264
- 126
I can't tell if the AMD has better AF, or if the green hue is causing a wash out effect. If that comparison actually proved better IQ on the AMD side, I wouldn't have to spend so much time playing around with the hue, brightness, and contrast on the images we capture from the high speed camera here in our lab (you know, to make the object captured pop out more). It'd save me a lot of time if things didn't work like that. :\
No, not FPS. Important things like component failure and life expectancy for example. These are hard to study, but so much more important than skipping some frames that you most likely will not notice. Other sites show tonality of cooler sounds, which matter if you get annoyed by coil whine and the like. Things like that which you will notice and really suffer from. You with me now?
njdevilsfan87
Platinum Member
- Apr 19, 2007
- 2,341
- 264
- 126
... and they look practically the same outside of small color differences. What am I looking for again? Improved IQ to the left? I don't see it. More MS to the left? I don't see it either.
I found myself thinking,
"Left side looks better. Wait nope, the right does now. Wait no uh..."
Or...
"Left side definitely has MS. Or wait, I think the right side has a more consistent but smaller MS. So the right side is worse. Wait no, I think it's actually smoother. No wait the screen just tore there, but the left didn't so the left side seems smoother again. Damnit the left just tore but the right didn't, I give up."
This is what your post said before you edited it:
Indeed, I guessed wrong. I figured the one on the right was the GTX680 because that side has better IQ, and it's a TWIMTBP title. To me it seemed like the left side had better looking Ambient Occlusion.
So, splain dat wun.
You're probably thinking AMD viral marketing, but it was actually just a mistake in my original post that I caught when editing the spoiler tag in.
The one on the left looked to me to have better AO, but I had a brain fart and said right on accident.
Keysplayr
Elite Member
- Jan 16, 2003
- 21,211
- 50
- 91
Think about it logically for a second. I originally said that I figured that the one on the right was the gtx680 because I thought it had better image quality and then went on to talk about how I noticed it in the ambient occlusion on the left side.
If it gets you excited thinking you caught me in some horrible lie more power too you, but it was just a typo. It happens.
This is a classic case of being influenced by knowing which was which. Once they were done blind he couldn't tell which card was which.
Red Hawk
Diamond Member
- Jan 1, 2011
- 3,266
- 169
- 106
Just a thought -- TechReport only tested the 7950 with boost under 12.11, correct? Back when the 7950 with boost came out, Anandtech did a little testing and found that the 7950 rather haphazardly switched between its base clock and boost clock (admittedly, it was not possible to precisely track the clock speed of the 7950, and they tested an original 7950 which had been flashed to use boost, although like with the 7970 GHz Edition there aren't supposed to have been any changes on the hardware side between the release 7950 and the boost upgrade). The issue might be related to the inconstant boost clock.
Most 7000 series cards don't use boost, just the current 7950 and the 7970 GHz edition. If the issue is boost-related, it would only effect those cards, and maybe not even the 7970 GHz Edition since its boost simply changes clock speeds, while the 7950's boost changes clock speed and voltage.
Most 7000 series cards don't use boost, just the current 7950 and the 7970 GHz edition. If the issue is boost-related, it would only effect those cards, and maybe not even the 7970 GHz Edition since its boost simply changes clock speeds, while the 7950's boost changes clock speed and voltage.
BrightCandle
Diamond Member
- Mar 15, 2007
- 4,762
- 0
- 76
If you listen to the tech report podcast you will hear they didn't just test boost, they tested a reference card as well. They saw the same problem.
Keysplayr
Elite Member
- Jan 16, 2003
- 21,211
- 50
- 91
Red Hawk
Diamond Member
- Jan 1, 2011
- 3,266
- 169
- 106
RussianSensation
Elite Member
- Sep 5, 2003
- 19,458
- 765
- 126
PCGames hardware posted new videos of Far Cry 3 to test micro-stutter
Far Cry 3 - GTX690 SLI vs. HD7970 GE CF
GTX690 SLI video as a gamer would see it on his monitor:
http://www.youtube.com/watch?feature=player_embedded&v=YaCbI8nx8Fo
HD7970 CF video as a gamer would see it on his monitor:
http://www.youtube.com/watch?feature=player_embedded&v=WXi_ra7Zr14
GTX690 SLI stutters more, unlike HardOCP's conclusion. Here are the 2 videos combined into slow motion (on the top left you can select HDD ON if it doesn't do it automatically):
http://videos.pcgameshardware.de/hdvideo/12558/Far-Cry-3-Mikroruckeln-in-Slow-Motion
How come SLI stuttered more in PCGamesHardware testing? You can't expect reviewers to test the same sections of a 20 hour game. Depending on where the game is tested, you can get different results. Just like we need to see 10-20 GPU reviews for FPS, now we are going to need 10-20 websites to make videos of all these games to draw more statistically accurate conclusions.
Nordic Hardware also did a similar comparison to TechReport. The video for BF3 was linked above but actually 680 technically stuttered more, although I couldn't even tell the difference in the video because the frame times are good enough for me even though GTX680 spiked to 30ms on several occasions. Did you guys notice those spikes in the video? I didn't.
BF3 - HD7970Ghz vs. GTX680
Here is a video too:
http://www.youtube.com/watch?feature=player_embedded&v=qR3ewLMbywY
HD7970Ghz stutters less and shockingly look at the frame rates at 0:30-0:32:
HD7970Ghz = 67 fps (+43% faster!)
GTX680 = 47 fps
Moral of the story:
- If you spend enough time, you can end up with unrealistically skewed results, either by accident or on purpose. In fairness, you can end up with different results based on a different section of a game / map and arrive at opposite conclusions from someone else who tested in a different level and for a different period of time (30 seconds vs. 1 min 30 seconds). Do we honestly believe HD7970Ghz is up to 43% faster in BF3? I never saw a single website review where HD7970Ghz absolutely crushed GTX680 in BF3 in FPS like that but Nordic Hardware video clearly shows it's not even close. Should we start claiming HD7970 is 30-40% faster than GTX680 in BF3 in many places? Guess what there is a video that "proves" it is. If HD7970Ghz crushes GTX680 in a certain part in a certain BF3 map at a certain resolution, is that a realistic expectation of the entire gaming experience in BF3? Not at all. That's the point, if you wanted to or by accident, you can show GTX680 getting leveled by an HD7970Ghz in BF3 in a 1 minute video because you accidentally picked an area where GTX680 tanks or vice versa. Chances are if you that sensitive to even 25-30 ms frame times in twitchy shooters, you are going to need to lower graphical settings in this game on both cards or go dual-GPUs for 60 fps minimum at all times.
- There is micro-stutter on all GPUs in all videogames it's just a matter of how much of it you can notice beyond a certain threshold that bothers you. This is different for every person. Different GPU architectures will stutter more in certain games/engines depending on how powerful the GPUs are and which manufacturer better optimized the drivers for that particular game / game engine.
Unless someone makes a GPU review with videos of 100-1000 games tested and statistically shows that 1 brand stutters more than the other by 50% or more, testing 1-5 games is not giving us the full picture. HardOCP concludes that GTX690 SLI stuttering less in FC3, PCgameshardware shows the opposite. By asking people their actual gaming experience with SLI/CF and checking many more reviews, most people tend to agree that SLI is less prone to stutter but it doesn't mean that in some games it won't stutter more than HD7970Ghz CF. You can still have games where GTX690/680 SLI is stuttering more than HD7970Ghz CF. Just like we wouldn't rely only on 1-2 reviews for FPS data, we should wait until more websites start using videos and seeing more and more games, not just 2012 games, but also 2013, 2011, 2010, 2009 games. Are there stuttering problems in some games on AMD cards? For sure there are and it's good AMD acknowledges this and TechReport started to investigate game testing with a different approach than FPS that focuses on smoothness. That doesn't suddenly mean if you get GTX680, it will stutter less than HD7970Ghz. It just depends on what games you play. We all know NV cards run faster in certain games and AMD cards run faster in other games. Just purchase the card that works better for the games you want to play based on 'general' consensus on which GPUs work better in that game, and just enjoy the game. If for example a gamer mainly plays Borderlands 2 or AC3, of course it's going to stutter way more on AMD cards. Also we know certain games on the PC run terribly to begin with like GTA IV, Rage, and just about all Bethesda titles. Is it news to people that games like Skyrim have stuttering issues when it's been a known problem for this engine?
If you sit there and spend all of your brain's capacity on actually trying to catch millisecond loading of each frame rendered and then you need a 120-240 Fps high speed camera to actually see it, is that how you play your games on your PC monitor? If you do that, you can find micro-stutter in a ton of games on AMD or NV GPUs and in many different parts of the game. That has been true for years but you probably didn't notice until someone told you to focus 100% on frame rendering and not playing the game. Micro-stutter happens on all GPUs, 680s, 680 SLI, 7970, etc. Just like tearing happens on all GPUs. And also, who plays games like Skyrim without Vsync at 100 fps and with no mods? Not only is that a screen tearing galore, but it's not representative of how the end user actually experiences the game. If you are running Skyrim with mods at 2560x1600, you are not thinking micro-stutter, but you are thinking boy what GPU do I need to get above 30 fps console frame rates so it's not a slideshow!
I welcome more reviewers doing actual video comparisons of games and especially without high speed cameras because that's how I get to see the game. If I can't notice micro-stutter on a PC monitor in real world motion, I don't really care that it's there at 240 FPS as I am never going to see it. The Nordic Hardware and PCGameshardware videos are actually more informative to me since I notice if the stutter will be evident or not.
Far Cry 3 - GTX690 SLI vs. HD7970 GE CF
GTX690 SLI video as a gamer would see it on his monitor:
http://www.youtube.com/watch?feature=player_embedded&v=YaCbI8nx8Fo
HD7970 CF video as a gamer would see it on his monitor:
http://www.youtube.com/watch?feature=player_embedded&v=WXi_ra7Zr14
GTX690 SLI stutters more, unlike HardOCP's conclusion. Here are the 2 videos combined into slow motion (on the top left you can select HDD ON if it doesn't do it automatically):
http://videos.pcgameshardware.de/hdvideo/12558/Far-Cry-3-Mikroruckeln-in-Slow-Motion
How come SLI stuttered more in PCGamesHardware testing? You can't expect reviewers to test the same sections of a 20 hour game. Depending on where the game is tested, you can get different results. Just like we need to see 10-20 GPU reviews for FPS, now we are going to need 10-20 websites to make videos of all these games to draw more statistically accurate conclusions.
Nordic Hardware also did a similar comparison to TechReport. The video for BF3 was linked above but actually 680 technically stuttered more, although I couldn't even tell the difference in the video because the frame times are good enough for me even though GTX680 spiked to 30ms on several occasions. Did you guys notice those spikes in the video? I didn't.
BF3 - HD7970Ghz vs. GTX680
Here is a video too:
http://www.youtube.com/watch?feature=player_embedded&v=qR3ewLMbywY
HD7970Ghz stutters less and shockingly look at the frame rates at 0:30-0:32:
HD7970Ghz = 67 fps (+43% faster!)
GTX680 = 47 fps
Moral of the story:
- If you spend enough time, you can end up with unrealistically skewed results, either by accident or on purpose. In fairness, you can end up with different results based on a different section of a game / map and arrive at opposite conclusions from someone else who tested in a different level and for a different period of time (30 seconds vs. 1 min 30 seconds). Do we honestly believe HD7970Ghz is up to 43% faster in BF3? I never saw a single website review where HD7970Ghz absolutely crushed GTX680 in BF3 in FPS like that but Nordic Hardware video clearly shows it's not even close. Should we start claiming HD7970 is 30-40% faster than GTX680 in BF3 in many places? Guess what there is a video that "proves" it is. If HD7970Ghz crushes GTX680 in a certain part in a certain BF3 map at a certain resolution, is that a realistic expectation of the entire gaming experience in BF3? Not at all. That's the point, if you wanted to or by accident, you can show GTX680 getting leveled by an HD7970Ghz in BF3 in a 1 minute video because you accidentally picked an area where GTX680 tanks or vice versa. Chances are if you that sensitive to even 25-30 ms frame times in twitchy shooters, you are going to need to lower graphical settings in this game on both cards or go dual-GPUs for 60 fps minimum at all times.
- There is micro-stutter on all GPUs in all videogames it's just a matter of how much of it you can notice beyond a certain threshold that bothers you. This is different for every person. Different GPU architectures will stutter more in certain games/engines depending on how powerful the GPUs are and which manufacturer better optimized the drivers for that particular game / game engine.
Unless someone makes a GPU review with videos of 100-1000 games tested and statistically shows that 1 brand stutters more than the other by 50% or more, testing 1-5 games is not giving us the full picture. HardOCP concludes that GTX690 SLI stuttering less in FC3, PCgameshardware shows the opposite. By asking people their actual gaming experience with SLI/CF and checking many more reviews, most people tend to agree that SLI is less prone to stutter but it doesn't mean that in some games it won't stutter more than HD7970Ghz CF. You can still have games where GTX690/680 SLI is stuttering more than HD7970Ghz CF. Just like we wouldn't rely only on 1-2 reviews for FPS data, we should wait until more websites start using videos and seeing more and more games, not just 2012 games, but also 2013, 2011, 2010, 2009 games. Are there stuttering problems in some games on AMD cards? For sure there are and it's good AMD acknowledges this and TechReport started to investigate game testing with a different approach than FPS that focuses on smoothness. That doesn't suddenly mean if you get GTX680, it will stutter less than HD7970Ghz. It just depends on what games you play. We all know NV cards run faster in certain games and AMD cards run faster in other games. Just purchase the card that works better for the games you want to play based on 'general' consensus on which GPUs work better in that game, and just enjoy the game. If for example a gamer mainly plays Borderlands 2 or AC3, of course it's going to stutter way more on AMD cards. Also we know certain games on the PC run terribly to begin with like GTA IV, Rage, and just about all Bethesda titles. Is it news to people that games like Skyrim have stuttering issues when it's been a known problem for this engine?
If you sit there and spend all of your brain's capacity on actually trying to catch millisecond loading of each frame rendered and then you need a 120-240 Fps high speed camera to actually see it, is that how you play your games on your PC monitor? If you do that, you can find micro-stutter in a ton of games on AMD or NV GPUs and in many different parts of the game. That has been true for years but you probably didn't notice until someone told you to focus 100% on frame rendering and not playing the game. Micro-stutter happens on all GPUs, 680s, 680 SLI, 7970, etc. Just like tearing happens on all GPUs. And also, who plays games like Skyrim without Vsync at 100 fps and with no mods? Not only is that a screen tearing galore, but it's not representative of how the end user actually experiences the game. If you are running Skyrim with mods at 2560x1600, you are not thinking micro-stutter, but you are thinking boy what GPU do I need to get above 30 fps console frame rates so it's not a slideshow!
I welcome more reviewers doing actual video comparisons of games and especially without high speed cameras because that's how I get to see the game. If I can't notice micro-stutter on a PC monitor in real world motion, I don't really care that it's there at 240 FPS as I am never going to see it. The Nordic Hardware and PCGameshardware videos are actually more informative to me since I notice if the stutter will be evident or not.
Last edited:
RussianSensation
Elite Member
- Sep 5, 2003
- 19,458
- 765
- 126
Bingo. You can test 4 games like Metro 2033, Sleeping Dogs, Dirt Showdown and Arma II and end up with HD7970Ghz outperforming GTX680 in micro-stutter. Alternatively, if you include games like BL2, Assassin's Creed 3 or Skyrim without mods (that has known issues on AMD cards stuttering without VSync enabled), and it shouldn't be a surprise that GTX680 would win. That's not exactly rocket science.
We already have conflicting results from videos where Nordic Hardware shows HD7970Ghz crushing GTX680 in BF3 in FPS and having lower stuttering, while TechReport has HD7950 trailing GTX660Ti in MOH:W based on the same engine. Similarly, HardOCP found less stuttering on 680 SLI than 7970Ghz CF in Far Cry 3 but PCgameshardware found more stuttering in 690 SLI vs. HD7970Ghz CF. ^_^
Will Robinson
Golden Member
- Dec 19, 2009
- 1,408
- 0
- 0
Very nice summation of the current state of play re" the stuttery AMD cards".
I think you have nicely put that in context although there's a certain NVDA devotee here that won't be pleased to see his "scandal/"investigation" so thoroughly refuted.
Bottom line....HD7970Ghz Edition spanks GTX680,both cards can be made to "stutter" according to the game and this whole witch hunt about some kind of "cheating drivers for more FPS" is debunked.
I think you have nicely put that in context although there's a certain NVDA devotee here that won't be pleased to see his "scandal/"investigation" so thoroughly refuted.
Bottom line....HD7970Ghz Edition spanks GTX680,both cards can be made to "stutter" according to the game and this whole witch hunt about some kind of "cheating drivers for more FPS" is debunked.
GotNoRice
Senior member
- Aug 14, 2000
- 329
- 5
- 81
The only reason the Geforce cards appear to stutter in that comparison is because they are running at an FPS below the monitor's refresh rate while the AMD setup is not. This is the conclusion the reviewers themselves came up with, as shown in the german text next to each video.
There is a specific issue at hand, seemingly caused by the latest betas that is causing frame latency issues on AMD hardware, and this isn't it.
Last edited:
Just a quick question and I apologize if its been hashed over the in the previous pages, but can youtube's 30fps (or is it 15?) show the MS issue in full light?
I've seen no evidence of this claim. Can you link to the tests that showed this? TR said that 12.8 was no better than 12.11b's.
BrightCandle
Diamond Member
- Mar 15, 2007
- 4,762
- 0
- 76
Exactly. TR themselves said in their report and in the podcast that they found no difference with 12.8 to 12.11. 12.11 isn't responsible for the results they got.
I personally saw no difference in stutter from the launch beta to the driver I was using 6 months later, but I never tested 12.8 so there is a chance this driver made it better, otherwise its likely from what I saw its been there since launch.
RussianSensation
Elite Member
- Sep 5, 2003
- 19,458
- 765
- 126
Ya and so do AMD's 7970GE cards because they don't even come close to 60 fps either. The main reason the 680s are stuttering even more than 7970s in that video is because they are running even slower in that area of the map than 7970s.
Source
That's the point I am making -- you can make 2 different videos in the same game with the same videocards and find places where HD7970CF stutters way less than 680 SLI and vice versa. This is why PCgameshardware and HardOCP have conflicting results.
You can do exactly the same with single GPUs.
Skyrim - Test Area 1
Skyrim - Test Area 2
Now imagine HardOCP tested Skyrim in Test Area 1. What would they conclude regarding HD7950 vs. 660Ti? Let's say AnandTech tested Area 2. Their conclusion would be totally different.
If you randomly pick 100 different test areas in Skyrim, what are the average results in terms of HD7950 vs. 660Ti in Skyrim in smoothness? You are telling me you are 100% confident that GTX660Ti won't stutter more in at least some of those 98 other areas of Skyrim not tested?
The BF3 Nordic Hardware review shows HD7970GE destroying GTX680 in 1 small area of the map, but it doesn't mean GTX680 is a bad card for BF3 overall. If a review just relied on Nordic Hardware's testing of BF3, what would you conclude that GTX680 is terrible for BF3? You would have.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter