3GB is a bottleneck for 1080p and we don't even need the hottest AAA titles for that. I've had several 3GB and 4GB cards (GTX 780, GTX 780 Ti, GTX 970, GTX 970 SLI, GTX 980, R9 280, R9 290 and R9 290 Crossfire) and the VRAM just isn't enough anymore even with 1080p60 when cranking those settings near to the maximum, even when the GPU itself could have enough performance.
GTX 1060 3GB was bottlenecked by the VRAM since it was released, the difference between shaders simply isn't the only thing which makes it slower. Dunno which game benchmarks you checked, but it's everything else than enough these days
Again TPU test results have
conclusively shown otherwise. You simply can not argue with the numbers. As it's been agreed that the shaders provide additional performance, then the impact of VRAM should b aptly demonstrated as one increases in resolution, the differnce between 1080p and 1440p shoud widen. It does not. Not only is it not affecting performance at 1080p, it is not affecting performance at 1440p.
Here's the TPU performance data for 1080p which shows a 6% difference between the 3 GB and 6 GB versions. . If , as you claim, "the difference between shaders simply isn't the only thing which makes it slower", then it inevitably and inarguably follows that the difference MUST be significantly larger at 1440p.
So what do we see at 1440p ? We see the same exact 6 % difference.
Performance is not impacted by VRAM until we get to 2160p where the performance difference uis 14%
Only 1 of 2 things can be true:
1. There is no significant performance difference between the 3 GB and 6GB model,. or
2. VRAM requirements do not change with increased resolution.
I don't think you will find much support for No. 2
As long as this argument has raged since the 6xx series but test results have continually proven forum postings to be exaggerated.:
With the 6xx....
Puget sound went at it trying to make a case for a 4HGB over a 2 GB, they could not. Look at the performance graphs
Similar video cards are often available in versions with more than one memory size. The GeForce GTX 680 is an example, and comes in both 2GB and 4GB variants. With computer components more is often better, but does doubling the memory on a video card like this actually help with game performance...
www.pugetsystems.com
"So, what can we glean from all of that? For one thing, with any single monitor the 2GB video card is plenty - even on the most demanding games and benchmarks out today. When you scale up to three high-res screens the games we tested were all still fine on 2GB, though maxing out some games’ settings at that resolution really needs more than a single GPU to keep smooth frame rates. With the right combination of high resolution and high detail, though, there is the potential for 2GB to be insufficient. For future games, or perhaps current games that were not tested in this article, you might be better off with a 4GB card if - and only if - you plan to run across a multi-screen configuration. "
With the 7xx....
-Alienbabeltech compared the 770 2GB and 4GB models and their results showed that the only toime they could detect a performance difference @ 5760 x 1080 between 2 Gb and 4 GB was when the game as unplayable anyway in the 15 to 25 fps range.
Do you need 4GB of ram? We tested EVGA's GTX 770 4GB versus Nvidia's GTX 770 2GB version, at 1920x1080, 2560x1600 and 5760x1080.
alienbabeltech.com
"We only start to see minimal differences at 5760×1080, and even so, there is rarely a frame or two difference. If we start to add even more AA, in most cases, the frame rates will drop to unplayable on both [ the 2 GB and 4 GB] cards ... This leaves five games out of 30 where a 4GB GTX 770 gives more than a 1 frame per second difference over the 2GB-equipped GTX 770. And one of them, Metro: Last Light still isn’t even quite a single frame difference. ,,, Of those five games, two of them are unplayable at 5760×1080 although in these cases .... There is one last thing to note with Max Payne 3: It would not normally allow one to set 4xAA at 5760×1080 with any 2GB card as it claims to require 2750MB. However, when we replaced the 4GB GTX 770 with the 2GB version, the game allowed the setting. And there were no slowdowns, stuttering, nor any performance differences that we could find between the two GTX 770s."
With the 9xx....
Guru 3D found that 1080p was more than adequate at 1080p
Remember the forums all full of outrage over the 970's 3.5 GB fake problem ?
It has been an interesting week with all the hype on the GeForce GTX 970 with what now is know as a crippled memory sub-system. Three weeks ago here on Guru3D.com users in our forums started to report...
www.guru3d.com
"Thing is, the quantifying fact is that nobody really has massive issues, dozens and dozens of media have tested the card with in-depth reviews like the ones here on my site. Replicating the stutters and stuff you see in some of the video's, well to date I have not been able to reproduce them unless you do crazy stuff, and I've been on this all weekend. .. Let me clearly state this, the GTX 970 is not an Ultra HD card, it has never been marketed as such and we never recommended even a GTX 980 for Ultra HD gaming either. So if you start looking at that resolution and zoom in, then of course you are bound to run into performance issues, but so does the GTX 980. These cards are still too weak for such a resolution combined with proper image quality settings. "
Is 4GB enough for a high-end GPU? We investigated and tested 15 titles to find out.
www.extremetech.com
"out of the fifteen games we tested, only four of could be forced to consume more than the 4GB of RAM. In every case, we had to use high-end settings at 4K to accomplish this. "
and at those settings ... the fps was at a point (hi 20s) that it would be hard to call them "playable". The point being ... by the time you got to a point here VRAM mattered, the card could npt deliver performance that one could call "playable"
Test data starts off point that way but finishes saying otherwise... but in ant case the 1060 3GB id faster than the 470 4GB even w/o considered that marked difference in OC ability.
The problem isn't that people are wrong, they have simply been misinformed. We have access to utilities that don't tell us what we think they do. The Max Payne test is the most telling ... the developer will not permit Max Payne to be loaded at 5760 ... with a 2 GB card installed. But if you fool it into doing to by installing with a 4 GB card, you can even use a utility to show you that it is using 3.75 MB. And yet removing that 4 GB card and replacing it with the 2 GB card and it plays at the same performance level and visual quality regardless of which card is installed.
No doubt there are game out there which have issues ... most of these have been attributed to poor console ports, other causes or reasons undetermined. Hitman is one game where VRAM does show an impact whiuch I can only attribute to VRAMYou can handpick your games and do things a certain way to "prove" any hypothesis you want... a common tactic at "contrarian" web sites. But the fact remains, we get the same performance in TPUs extensive game test suite at 1440p as we do at 1080p. That destroys any claim that 3 GB is going to be a problem in the majority of instances.... And when we do see reductions from say 95 to 95 or 115 to 108 ot 87 to 77 ... of what impact is this when the great majority of screens out there are 60 - 75 Hz ?both RX cards are slower outta the box and the 1060 has 40% more overclocking headroom than the RX cards have.
