GTX 1060 or RX480? OR a GTX 970 for Cheap.

LinkedPc

Honorable
Jul 1, 2013
51
0
10,630
Looking for a upgrade, between a 1060 and a 480 what would be better for me in the long run? (next 4-5 years) Also I am wondering if it would be worth it to buy a cheap gtx 970 and OC it? I'm seeing new/barley used ones for as low as $150... Would that be a viable option? What do you guys think? NOTE: I game in 1080p as of now, maybe looking to upgrade in the future. Also for workstation purposes what would be the "jack of all trades" choice? (ex. premiere pro, after effects, photoshop) Thanks.
 
Solution
As you can see below, the reference 1060 is 10% faster than the reference 480 (97/88) and 17% faster than the reference 970 (97/83). If you overclock.... from a gaming perspective:

perfrel_1920_1080.png


TPUs test of an AIB 480 showed the ability to overclock 6% over reference
TPUs test of an AIB 1060 showed the ability to overclock 18% over reference (24% faster then AIB 480)
TPUs test of an AIB 970 showed the ability to overclock 17% over reference (5% faster than AIB 480

The AIB cards generally fall within 1-2% of one another.

For your workstation usage (video editing with Premiere), you will want a CUDA capable card which is an nVidia only thing.

After that, it's...
I think you would want an Nvidia card for workstation purposes. (Maybe someone else can comment on this).

As for gaming, the VRAM of the 970 is going to become a problem (It already is a problem in some games), so you're going to be better off with a 1060 or RX 480. While the 1060 is much better for DX11 games, the RX 480 is very promising for games that have implemented DX12/Vulcan features correctly. So in the long run, looking at where games are going, I think an RX 480 would be better since Nvidia is also known for gimping their old cards.
 
As you can see below, the reference 1060 is 10% faster than the reference 480 (97/88) and 17% faster than the reference 970 (97/83). If you overclock.... from a gaming perspective:

perfrel_1920_1080.png


TPUs test of an AIB 480 showed the ability to overclock 6% over reference
TPUs test of an AIB 1060 showed the ability to overclock 18% over reference (24% faster then AIB 480)
TPUs test of an AIB 970 showed the ability to overclock 17% over reference (5% faster than AIB 480

The AIB cards generally fall within 1-2% of one another.

For your workstation usage (video editing with Premiere), you will want a CUDA capable card which is an nVidia only thing.

After that, it's just a matter of balancing budget and performance .... but with $150 for the 970, that seems quite attractive if budget is an issue. But if you can manage the budget,the 1060 is clearly the better card. Whether it is worth 3 ti.mes the price to you is another story

The VRAM broohaha about the 970s is completely bogus. You can create an issue if you try really really hard, but since the **story** broke, numerous web sites have been trying to duplicate the problem w/o success. The 970 has no issues whatsoever up to 1440p and., at that point, so does the 980. It's all quite clearly laid out on dozens of sites ... here's one of them:

http://www.guru3d.com/news-story/middle-earth-shadow-of-mordor-geforce-gtx-970-vram-stress-test.html

We understand there have been some questions about how the GTX 970 will perform when it accesses the 0.5GB memory segment. The best way to test that is to look at game performance. Compare a GTX 980 to a 970 on a game that uses less than 3.5GB. Then turn up the settings so the game needs more than 3.5GB and compare 980 and 970 performance again.

On GTX 980, Shadows of Mordor drops about 24% on GTX 980 and 25% on GTX 970, a 1% difference. On Battlefield 4, the drop is 47% on GTX 980 and 50% on GTX 970, a 3% difference. On CoD: AW, the drop is 41% on GTX 980 and 44% on GTX 970, a 3% difference. As you can see, there is very little change in the performance of the GTX 970 relative to GTX 980 on these games when it is using the 0.5GB segment.

After some internal testing here over the weekend we could quite honestly not really reproduce stutters or weird issues other than the normal stuff once you run out of graphics memory. Once you run out of ~3.5 GB memory or on the ~4GB GTX 980 slowdowns or weird behavior can occur, but that goes with any graphics card that runs out of video memory.

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.

Overall you will have a hard time pushing any card over 3.5 GB of graphics memory usage with any game unless you do some freaky stuff. The ones that do pass 3.5 GB mostly are poor console ports or situations where you game in Ultra HD or DSR Ultra HD rendering. In that situation I cannot guarantee that your overall experience will be trouble free, however we have a hard time detecting and replicating the stuttering issues some people have mentioned.

In short, the 970 performs no differently than the 980



 
Solution
GTX970 cons:

1) 3.5GB of fast memory, vs 6GB for GTX1060

2) GTX1060 up to 1.6X faster in VR (on top of any performance difference already)

3) GTX1060 has asynchronous optimizations which DX12 can utilize. These do two important things:

a) can boost FPS slightly, and
b) can prevent STUTTERING in many cases.

The last point is important. The difficulty in switching between compute and other tasks often results in some stuttering. This will help avoid those issues.

4) GTX1060 also improves a few other things:
- HDR support
- higher DP/HDMI bandwidth
- better HEVC encode/decode

There are some minor things I've missed.
 


If I really want to I can get a 1060 or a 480, I've had people saying the RX480 would be better for me in the long run (future games = DX12) but obviously the 1060 is better for right now. (DX11) What are your thoughts on this?
 
Dx12 is damn impressive on amd hardware, agree. But for pure performance the 1060 for majority of games. My only experience is vulkan on Doom which takes the 390 up to 1060 level of fps, but this isn't going to apply for all games and the 1060 could still outperform the 480 even with the dx12 advantage. Still my vote to 1060
 


I'm not buying it.... while I keep holding out hope that AMD will again be competitive, I have been too many times disappointed...

-They haven't been competitive at the top tiers in 3 generations
-The owned the R9 x80 versus GTX x60 niche over that period but lost bad this time
-Everything about positive about AMD started with "will be" .... we ever seem to get to "is"
-The claims that [insert new AMD technology here] is going to change everything never seem to materialize ... Mantle didn't do anything, HBM didn't do anything. Don't get me wrong... HBM is coming but HBM1 (w/ its 4 GB limit) was a wasted effort.
-And that's how I feel about DX12.... Using DX12 is all about drivers.... I expect that the drivers will be here when more games are here. In addition, all we see in most game benchmarks are stock settings ... the 1 or 2 fps advantage the 480 is now seeing w/ DX12 are more than offset by the the green cards having 3 times as much of a performance increase gained from OCing.

So as much as I wanted AMD to "take this one", I think we will have to wait for a 480x for a possible win.

 


Many functions in Premiere will perform equally on either card as those functions are CPU bound ... but some benefit greatly from CUDA ... Adobe After Effects also benefits from CUDA, an nvidia only technology

https://www.pugetsystems.com/recommended/Recommended-Systems-for-Adobe-Premiere-Pro-143/Hardware-Recommendations
https://www.studio1productions.com/Articles/PremiereCS5.htm

As fior the rest of the specs.... I'd have no problem ignoring most of these, but let's say for a moment that fps was equal rather than 24% apart, the ...

480 is noisier
480 uses more power
480 runs hotter
480 can't overclock more than single digits
480 can't use PhysX
480 can't use ShadowPlay
480 w/ Freesync can't provide ULMB

As for the 8GB / 6 GB thing

https://www.techpowerup.com/reviews/ASUS/RX_480_STRIX_OC/28.html

The GTX 1060 is the better alternative in my opinion if the card is priced much higher, though, close to $300, as it provides better noise and thermals and higher performance; the difference between 6 GB and 8 GB has no effect on 1080p gaming, not even in the most demanding titles.

The same holds for 1440p.
http://www.extremetech.com/gaming/213069-is-4gb-of-vram-enough-amds-fury-x-faces-off-with-nvidias-gtx-980-ti-titan-x

If moving later to 4k, the 480 8GB in CF is a better choice than the 1060 at 4k.... problem is, two of these are $500+ and the 1070 is still a bit faster at stock settings, way faster when both cards are overclocked and the 1070 is quite a bit cheaper.

But again, with specific reference to this topic, the OP will be using Adobe Premiere which means he's better served with an nVidia card


 
Thanks to everyone for your opinions, I have asked other communities as well and I think I will wait and get a GTX 1060 or if I can find cheap GTX 970 deal I will go with that. Thanks for all the feedback and opinions!
 


AMD Polaris hardware (including rx480) and Windows 10 software are tightly integrated, so going forward dramatic graphics improvements are coming. DirectX 12 provides deeper access to the GPU than DirectX 11, meaning less overhead and better performance. It also enables a next-gen SLI effect, where two graphics cards are used at the same time. DirectX 12 loosens this up, letting the resources of two completely different cards be used simultaneously.

DirectX 12 enables new shader and rendering techniques for upgrading the quality of reflections, fog, smoke and lighting. It allows calculations to a per-pixel level. Not only does this make environmental effects look better, it make player interactions with things like smoke/particles easier to model.
 
Yes... but relevance ? Drivers continue to be revised as we have moved from DX10 to DX 11 to DX 12 ... and will continue to do so. Optimizing your drivers for any one of those means it is less optimized for the others. So what do you think is the better path ? As new games come out, both sides revise their drivers for better performance in those games, while effort is made not to decrease performance in others. Optimizing for DX11 which represents the great majority of the games on the market or optimizing for DX12 so you can uses DX12 performance as a sales pitch ?

But in the end, where's the significance to the audience here ? Having a 1-2 fps advantage in a small number of games (while 10% slower overall) and having that advantage immediately disappear once Afterburner is opened ?

Rise of the Tomb Raider (DX12)
1060 = 70.8 fls x 1.18 = 83.5
480 = 71.2 fps x 1.06 = 75.4

Hitman
"The game's DirectX 12 implementation, however, is too riddled with bugs to be integrated into our test bench for now. In this test, we're testing the game in DirectX 11 mode"

So out of 16 games ... TPU was able to test only 2 in DX12. One of the two games (50%) were to buggy to use, leaving us with just 6% of games on which it was a factor. If tasked with writing those drivers, my thought is concentrate more on the 94% of games for now and focus on DX12 as games start to enter the market.. Like a horse race, "going out early" is a strategy that has advantages as well as disadvantages. So while it is promising that AMD was able to improve performance with DX12, it still, so far, hasn't been able to bring them home a win.

I see you have read the MS marketing material ... but I have yet to see a review where it states ... "wow, the fog clearly looks better" ... Been hearing for years that "PhysX doesn't matter" and those are the kinds of features that PhysX added .... I heard the same thing bandied about wit the GTX 4xx series ... "you could fry an egg on it" ... and yet when the tables turned top where they now have been for the last few generations, all of a sudden it doesn't matter anymore. If a feature is important, it should matter all the time and not suddenly become meaningless once the competition has the advantage.

DX11.1 (Win8) was going to bring all sorts of wonderful things to the table and the claim was... everything will be faster in Win8 ... it wasn't.

As we can see here .... based upon the then current implementations, the biggest improvements where the fps increases seen on RotTR due to easing CPU burdens and spreading over multiple threads... but it still can't catch the 1060.

http://www.hardocp.com/article/2016/04/19/dx11_vs_dx12_intel_cpu_scaling_gaming_framerate/1

What we see is that DX12's advantages, like mantle, have a great impact on weaker CPUs and on lower resolutions, than on higher end ones.

Just like 11.1, we see MS's claims for DX12 not exactly "panning out", especially on high end hardware ... as for the claimed visual improvements ... no one has been able to actually see any significant difference. And AMD themselves is not singing any tunes about visual quality but state that DX12 is all about performance.

http://www.pcgamesn.com/ashes-of-the-singularity/how-different-does-ashes-of-the-singularity-look-in-dx11-and-dx12-err-not-very

Reading AMD's own sales pitch on DirectX 12 .... they give a very clear message that the improvements of 12 over 11 are performance-based

Over time, I expect DX12 to have a potentially significant impact, that time is not yet here. It does give developers a lot of tools ... the question, in these days or quick console ports, how many will bother to invest the time, money and effort in taking advantage of those tools.

Assuming that does take place ..... yes, AMD's strategy was to jump outta the gate and take the early lead in DX12 performance while nVidia focused their efforts elsewhere.... but the horse that has the lead at the quarter pole doesn't always cross the finish line 1st.

This is another case of "Deja Vu all over again". There was the same hype about Mantle and the claimed world changing benefits never materialized .. there was the same hoopla about DX 11.1 and the claimed world changing benefits never materialized. Here we are again ... on the visual improvements, we see articles titled "How different do games look in DX11 and DX12? Err... not very" ... on the performance side, yes we see improvements but not enough to erase nVidia 3x advantage in overclocking.

Will nVidia fall flat on their face with DX12 optimizations once DX12 is implemented in more games ? They might ... but should we base our purchasing decisions today on what might happen in the future .. especially when we've been disappointed when had our hopes up so many times before.

So until such time as these visual improvements are actually realized and until the RX cards can perform significantly better than the GTX cards such as to offset the noise, power, heat and feature disadvantages ... it's hard to see them coming out of a title bout with a win.

 

TRENDING THREADS