• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Need Help with picking gpu.

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
The best thing to gain from the information here, OP, is how long do you plan to own this card?

6 months to 1 year? 1060 will be about 2-5% faster in most of the games in that time compared to 480 (but not your 2 target games--480 will be better). After about a year, when nVidia support and performance drops off a cliff for the 1060, you can put up another $200-250 for their updated model.

The simple fact, based on well, facts and known history, is that the 480 will be beating the 1060 in every benchmark within a year's time, and will be wiping the floor with it in 2 year's time. You strike me as someone--moving from PS4 to your first PC--that isn't interested in dropping $200-300 every year to maintain/improve by about 10% performance.

Why not spend $250 now on a card that actually improves by that same 10-15% performance over 3 or 4 years without having to buy another card?

There is a reason that those encouraging you to get the 1060 think that you shouldn't worry about the future: Only now matters because if you are an nVidia user, you expect to replace that card in 6 months or a year, anyway. They want you to dismiss future considerations because the history of nVidia cards aging really is not favorable to actual consumers that are concerned with value propositions.
 
@ Zinfamous
"The simple fact, based on well, facts and known history, is that the 480 will be beating the 1060 in every benchmark within a year's time, "

I have a rx490 of your choice that says you'll be wrong. If i'm wrong you buy me a 1080 of my choice.

PM me if you want to put your money where your mouth is.
ANd if any of you guys that "liked" his post want some action , PM me also.

By the way, I'm fairly sure the OP bought a 1060 2 days ago.
 
I doubt every benchmark, but probably majority of them.

I have a rx490 of your choice that says you'll be wrong. If i'm wrong you buy me a 1080 of my choice.

There is no RX 490 so not sure what exactly you'd be buying.

You won't want a 1080 once Volta comes out anyway though, will be old news like the 980 Ti.
By the way, I'm fairly sure the OP bought a 1060 2 days ago.

Must have missed that post, guess this whole thing should have been closed then.
 
@ Zinfamous
"The simple fact, based on well, facts and known history, is that the 480 will be beating the 1060 in every benchmark within a year's time, "

I have a rx490 of your choice that says you'll be wrong. If i'm wrong you buy me a 1080 of my choice.

PM me if you want to put your money where your mouth is.
ANd if any of you guys that "liked" his post want some action , PM me also.

By the way, I'm fairly sure the OP bought a 1060 2 days ago.

lol, I don't make stupid bets. 😀
 
I doubt every benchmark, but probably majority of them.



There is no RX 490 so not sure what exactly you'd be buying.

You won't want a 1080 once Volta comes out anyway though, will be old news like the 980 Ti.


Must have missed that post, guess this whole thing should have been closed then.

It would take a year's time, so there would be a 490 by then (one would hope 😀), and dude has a 960--I think that 1080 would make him very happy. Beating every benchmark within one year is hyperbole on my part--I think winning in overall performance is likely in that time, probably not certain, but 1060 is highly likely to fall off a cliff within 2 years.
 
I don't think you can go wrong with either the RX 480 or the GTX 1060 6GB.. Both cards are really great.. Some advantages that the GTX 1060 has are less power usage, and better hardware acceleration for HEVC decoding whilst offering similar performance to the RX 480, or slightly more or less depending on the game.

I'm not convinced that the RX 480 will necessarily become more dominant over time. Just because a trend has existed, doesn't mean it will always exist. Pascal will be a tougher nut to crack as it has eliminated or dramatically reduced many of the weaknesses that Maxwell had, whilst shoring up the strengths.
 
He is only on a normal 1080p monitor and the 1070 would be overkill for that really.
I disagree. I am on 1080P and the 1080 I own still makes a HUGE difference compared to my 980 that it replaced. A 1070 in a 1080p system is not overkill. Period. It will still provide more FPS and a better gaming experience because of it. My 980 at 1080P couldn't run some of my games completely maxed out at 60 fps. My 1080 does it with room to spare. A 1070 would be worth it for the performance gained over the 1060 at 1080p.
 
Last edited:
1070 is the way to go. Big leap over the 480/1060. I love mine. It makes me smile.

I agree that 1070 is the current best card out there with the new nodes that hits that sweet spot of no worries 1080p60 performance/very high 1440p(60-75?) that should make most owners smile. Despite whatever limitations Pascal has with DX12 compared to Polaris, 1070 will be able to brute force its way through those for the next 3 or 4 years, on top of having no VRAM issues to consider in that time.

It's finally coming down to its actual MSRP or very close to it, which really makes it a no-brainer with the recent extortionate prices of the 480 8gb/1060 6gb that have been out there (though less so for the 1060). While those prices are coming down as well, it really makes no sense for anyone (outside of mining) to consider an AIB RX480 @ ~$300 if they can also find the magic 1070 @$370-390.

I still think a 480 (8gb) @ $240-260 or 1060(6gb) @ $250-270 is the best choice for the mainstream, but those aren't common enough.

1070 really is going to own that tier for the next 6+ months because AMD has dropped the ball on 490 (While their slower release with Vega was long planned and no mystery, most people should now accept that this is really going to be a Q2 2017 release. That's pretty terrible for them, imo. early Q1 2017 really should be their target, but it's no mystery what they mean when they mention 1H 2017)
 
The best thing to gain from the information here, OP, is how long do you plan to own this card?

6 months to 1 year? 1060 will be about 2-5% faster in most of the games in that time compared to 480 (but not your 2 target games--480 will be better). After about a year, when nVidia support and performance drops off a cliff for the 1060, you can put up another $200-250 for their updated model.

The simple fact, based on well, facts and known history, is that the 480 will be beating the 1060 in every benchmark within a year's time, and will be wiping the floor with it in 2 year's time. You strike me as someone--moving from PS4 to your first PC--that isn't interested in dropping $200-300 every year to maintain/improve by about 10% performance.

Why not spend $250 now on a card that actually improves by that same 10-15% performance over 3 or 4 years without having to buy another card?

There is a reason that those encouraging you to get the 1060 think that you shouldn't worry about the future: Only now matters because if you are an nVidia user, you expect to replace that card in 6 months or a year, anyway. They want you to dismiss future considerations because the history of nVidia cards aging really is not favorable to actual consumers that are concerned with value propositions.

Since you can predict the future, how will the cards compare to each other in VR in 3 or 4 years? Will AMD still have a "premium" VR experience?
 
Since you can predict the future, how will the cards compare to each other in VR in 3 or 4 years? Will AMD still have a "premium" VR experience?

That's one of the predictions from AMD fans that fell massively short. Supposedly AMD was going to dominate VR because of the ACEs and much faster ATW and blah blah blah. But looking at the benchmarks that I've seen, it seems that NVidia is consistently ahead of AMD when it comes to VR. And it's not just Pascal, but even Maxwell!
 
That's one of the predictions from AMD fans that fell massively short. Supposedly AMD was going to dominate VR because of the ACEs and much faster ATW and blah blah blah. But looking at the benchmarks that I've seen, it seems that NVidia is consistently ahead of AMD when it comes to VR. And it's not just Pascal, but even Maxwell!

Because none of the VR games are full games but just early access unoptimized tech demos.

Or can you explain why Nvidia is falling well shy of its 3x VR performance over Maxwell claims when the 980 Ti is consistently beating the 1070?
 
Because none of the VR games are full games but just early access unoptimized tech demos.

Or can you explain why Nvidia is falling well shy of its 3x VR performance over Maxwell claims when the 980 Ti is consistently beating the 1070?

The 3x claim of 1080 over 980 Ti is when SMP is factored in. SMP allows the GPU to do a ton less work to render the contents displayed in both eyes, leading to a significant speed up. Once SMP is implemented in VR games, you should see a dramatic speedup for Pascal GPUs over Maxwell GPUs.

I suggest studying NVIDIA's whitepaper on Pascal to get more details:

http://international.download.nvidi...al/pdfs/GeForce_GTX_1080_Whitepaper_FINAL.pdf
 
The 3x claim of 1080 over 980 Ti is when SMP is factored in. SMP allows the GPU to do a ton less work to render the contents displayed in both eyes, leading to a significant speed up. Once SMP is implemented in VR games, you should see a dramatic speedup for Pascal GPUs over Maxwell GPUs.

I suggest studying NVIDIA's whitepaper on Pascal to get more details:

http://international.download.nvidi...al/pdfs/GeForce_GTX_1080_Whitepaper_FINAL.pdf

Sorry I forgot the whitepapers being required reading before purchasing, they don't seem to highlight that url from their hompage:

LXQAdLq.png


What I'm trying to say is it's silly to claim Nvidia is so much better at VR than AMD when neither is currently being used properly. Proper VR games will use a low level API and make much better use of all hardware. Right now the only games tested are ones that aren't optimized for either group, or developed in conjunction with one of them, neither of which shows how release performance will be.

Do you agree that these games don't fully support either AMD or Nvidia's hardware features?
 
Because none of the VR games are full games but just early access unoptimized tech demos.
So what?

Predicting Nvidia having a performance advantage some number of years in the future is the same thing as predicting AMD is going to have a performance advantage some number of years in the future.

Or to put it in your words:
What I'm trying to say is it's silly to claim AMD is so much better at DX12 than Nvidia when neither is currently being used properly. Proper DX12 games will use a low level API and make much better use of all hardware. Right now the only games tested are ones that aren't optimized for either group, or developed in conjunction with one of them, neither of which shows how release performance will be.
 
Back
Top