AMD Polaris architecture – GCN 4.0

[Kaapstad]

yes but you have settings that gives you a very high quality image at first sight, and you have settings that you need a magnifying glass and 30sec of staring to spot a difference.
as many others explained, you can get max image quality or extremly close to it , without throwing away performance in useless settings.
therefor you dont need 300% increase of flagship perf to have playable 4k, a 2x390 would be plenty enough.

Every ones visual experience and priorities are different.

You can not tell me what I don't need any more than I can tell you what you do need.

What is can do is state what hardware is needed to use all the settings in a game to the max, that does not mean that you or any other end user should use those settings.

 

[bru]

Why is 8x AA maxed out though, why not 16x aa, or 32xaa, or 32x aa + transparency AA and whatever else.

I didn't mention any set amount of AA.

Maxed used to mean highest main settings in the game, nothing more or less, 8xaa isn't always an option in games.

This is what I would suggest is maxed, whatever is the absolute highest settings the game developer has allowed, even if they actually look worse, but use more power to create.

 

[Seanspeed]

Vega most likely will be 300MM2 to 400MM2 if Polaris is well under 300MM2.

If AMD have some insider knowledge that Nvidia aren't going bigger than that, then sure.

Otherwise, that's leaving too much on the plate. Playing the budget GPU game hasn't worked out for them so far and there's no reason to think it will work out for them again. If anything, the 980Ti vs Fury X debacle only further increased Nvidia's market share among the enthusiast market.

 

[drunkenmaster]

If AMD have some insider knowledge that Nvidia aren't going bigger than that, then sure.

Otherwise, that's leaving too much on the plate. Playing the budget GPU game hasn't worked out for them so far and there's no reason to think it will work out for them again. If anything, the 980Ti vs Fury X debacle only further increased Nvidia's market share among the enthusiast market.

This couldn't be more wrong at all. The two generations AMD has made huge GPUs that weren't aiming to hit the best price/performance, 2900xt and the 7970 through Fury era are the periods AMD has been their least competitive in terms of profit/market share.

AMDs absolutely by a mile best period was 4870/5870 in terms of market share, competitiveness pricing. The 4870 was a ridiculously small chip, ultra efficient and incredibly well priced and it brought in a huge amount of sales and money.

The fury x vs 980ti debacle as you put it was AMD going all out with an ultra enthusiast and absolutely anything but budget chip. Fury X costs considerably more than 980ti to make. It is in every way an ultra premium chip using the latest tech, the latest most expensive production, packaging and memory. In every way it's a technological marvel that comes at a heavy price tag.

 

[alijayden583]

complete information

 

[andybird123]

The fury x vs 980ti debacle as you put it was AMD going all out with an ultra enthusiast and absolutely anything but budget chip. Fury X costs considerably more than 980ti to make. It is in every way an ultra premium chip using the latest tech, the latest most expensive production, packaging and memory. In every way it's a technological marvel that comes at a heavy price tag.

Just a shame that it was lopsided with not enough ROPs and it didnt get vaguely competitive in games for over 6 months, when most everyone in the market for a high end gpu had already bought a 980ti

 

[JediFragger]

If AMD have some insider knowledge that Nvidia aren't going bigger than that, then sure.

Otherwise, that's leaving too much on the plate. Playing the budget GPU game hasn't worked out for them so far and there's no reason to think it will work out for them again. If anything, the 980Ti vs Fury X debacle only further increased Nvidia's market share among the enthusiast market.

Wouldn't suprise me if the Ti outsold the FuryX by 1000 to 1 tbh

 

[CAT-THE-FIFTH]

If AMD have some insider knowledge that Nvidia aren't going bigger than that, then sure.

Otherwise, that's leaving too much on the plate. Playing the budget GPU game hasn't worked out for them so far and there's no reason to think it will work out for them again. If anything, the 980Ti vs Fury X debacle only further increased Nvidia's market share among the enthusiast market.

Nothing to do with that - look at all the AMD "large" GPUs on a new node - they are between 300MM2 to 400MM2 in size. Its about manufacturability - why do think the GP100 is being prioritised for commercial customers who will pay many times the amount even a Titan costs?

This is the latest info:

http://www.bitsandchips.it/52-english-news/6916-tesla-consumer-cards-only-during-1q17

Nvidia will just about be able to supply enough Tesla cards during Q1 2017.

 

[muziqaz]

Wouldn't suprise me if the Ti outsold the FuryX by 1000 to 1 tbh

yet they still managed to grab back some of the marketshare in past few quarters

If AMD has a killer product for mid range where nvidia cannot match them, AMD will be in much better position that they are now. From what we heard by now it seems AMD is targeting exactly that. very good performance with very low power usage and very well priced.

 

[AlamoX]

yet they still managed to grab back some of the marketshare in past few quarters

If AMD has a killer product for mid range where nvidia cannot match them, AMD will be in much better position that they are now. From what we heard by now it seems AMD is targeting exactly that. very good performance with very low power usage and very well priced.

the 390 is a killer product in mid range, double the memory of the 970, and better performer in the vast majority of games, and the 970 still sells better.
imagine if the 970 had 8Go and still competing with a 290 4Go, that would make it a no brainer for ppl, but yet here is not the case, AMD can have the better tech and the better product and still wouldn't sell better than Nvidia, thats just how ppl perceive things, for AMD to change that you would need a year or 2 of benchmark stamping on nvidia, especialy at high end.

 

[Seanspeed]

I think the 'mentality' arises when you reach a point that returns have diminished so much that you can no longer see the difference. Especially when Kaap starts talking about wanting maximum AA at 4K as in some other instances.

No-one would argue that no AA isn't worse than 2xAA and that therefore you've maxed out with no AA. But when you get to the levels that Kaap frequently talks about it's like someone tying lead weights to their belt and saying you need stronger leg muscles to run a half-marathon. People just don't see that this adds anything, which is why they stop accepting undiscernible extra levels as being "maxing something out".

What I wont ever understand is the psychology behind it. Some people seem to have this compulsive or dependent *need* to max something out, no matter what. Like it pains them to have to lower a setting. It seems a bit unhealthy. And expensive.

There's plenty of better examples than AA, too. Some games just throw an inordinate amount of options at users. Which is great. But for people who have this need to 'max' something out, it must be torture. Especially when they add in performance-torturing options that have little to no discernible difference, like higher resolution god rays in Fallout 4 and whatnot.

And then you have questions like, "Does resolution count as a graphics setting?". What about tweakable ini files where you can go nearly unlimited on certain options and overload things til the game wont even start you're asking so much of the GPU. What about resolution scaling, which is effectively just an in-game downsampling resolution slider.

Options are there to tailor the demands to varying system setups and for people to balance graphics vs performance the way they want to. That's all they are meant for. One game can throw in options that could cripple a 4x980Ti setup. Another developer might only give a small handful of options, yet have the same base graphics and the game would run well on a single GTX770. In the end, it's all fairly arbitrary.

What I really appreciate is the Nvidia guides that are put out for graphics options in many AAA games. They go to comprehensive levels to show and analyze the differences between each and every option on the board and how much they impact performance. I can often read through it and then have a very good idea of what settings to use, or at least what settings I can afford to lose, before I ever even start up a game. It also does a good job of showing that many setting differences are basically useless, adding essentially nothing yet still costs frames. Of course for the 'max settings' addicts, they'll still chase those pointless settings and they'll buy whatever hardware they need to in order to use those meaningless options that add nothing to the experience.

It's crazy.

 

[Seanspeed]

the 390 is a killer product in mid range, double the memory of the 970, and better performer in the vast majority of games, and the 970 still sells better.

It took quite a while before AMD's drivers got the 290/390 on level with the 970.

And it's certainly not a 'vast' majority of games it performs better in. The balance has shifted towards it over the past year, but the 970 still does better in quite a few titles. Both are entirely worthy options and I wouldn't try and make anybody feel bad whichever they bought.

But you're right that the 390 probably doesn't sell as well as it should given its price/performance. And that AMD have an image problem they need to get over before the status quo can be affected significantly. That said, a really clear Day 1 debut win for a card WOULD have a big impact. In the end, users just want the best performance and if AMD was the clear way to go, I dont think for a second that word wouldn't spread that 'x' card was the best option. But it has to be Day 1. They cant afford to put out a card that only 'wins out' after a year of driver development. By then, the damage will have been done.

 

[andybird123]

.

I wouldnt describe myself as a max settings addict, however even with a heavily overclocked 980ti (and lacklustre multi-gpu support in the titles i lve wanted to play) at 1440/144hz, i am having to turn down most settings by a notch or two to get sustained 90fps minimums.

I would love to have everything on high with 4xAA, that would be my sweetspot, but i often have to go with 2xAA with FXAA or some other compromise (if only more games would support SMAA which seems to give a better balance than MSAA)

 

[Seanspeed]

I wouldnt describe myself as a max settings addict, however even with a heavily overclocked 980ti (and lacklustre multi-gpu support in the titles i lve wanted to play) at 1440/144hz, i am having to turn down most settings by a notch or two to get sustained 90fps minimums.

I would love to have everything on high with 4xAA, that would be my sweetspot, but i often have to go with 2xAA with FXAA or some other compromise (if only more games would support SMAA which seems to give a better balance than MSAA)

That's all entirely reasonable. You're certainly not the sort of person I'm talking about.

And having a 144hz monitor definitely makes it so that compromises are almost always going to be necessary to take advantage of the high framerates possible.

 

[Overlag]

Wouldn't suprise me if the Ti outsold the FuryX by 1000 to 1 tbh

Not sure if outsold would be right... But the amount of free samples Nvidia give to people like Linus, Jayztwocents and other top youtube channels the "sales" must look good. No wonder all these channels use Nvidia cards, because they got them all free (nothing against the people, more Nvidia).

Its almost as bad as when Intel sucked, but still had massive lead over AMD because they paid off the whole market to keep using them etc.

 

[Mei]

it would be dumb not to
many will buy what they see youtubers and twitch streamers using
maybe they cant afford the same card but they can afford the same brand

 

[CAT-THE-FIFTH]

the 390 is a killer product in mid range, double the memory of the 970, and better performer in the vast majority of games, and the 970 still sells better.
imagine if the 970 had 8Go and still competing with a 290 4Go, that would make it a no brainer for ppl, but yet here is not the case, AMD can have the better tech and the better product and still wouldn't sell better than Nvidia, thats just how ppl perceive things, for AMD to change that you would need a year or 2 of benchmark stamping on nvidia, especialy at high end.

The launch was rubbish just like with Fury. If AMD had not skimped on the reference cooler and had decent third party cards out,and not allowed Nvidia to troll them,then things would be different.

Look at what a simple rebranding to R9 390/390X did!

This is why AMD needs a perfect launch like they did with the HD4000 and HD5000 series when Polaris is released.

 

[LoadsaMoney]

It better be a Godamn monster tbh, we need it to clean the Titan's clock and bring nVidia down a peg or 9

I can't see it myself.

 

[Mei]

yeh i dont see that happening
but if they run cooler and can overclock without watercooling maybe there's some hidden goodies
i rly think AMD should change the way their cards boost too
most gamers dont tinker!

 

[Kaapstad]

It better be a Godamn monster tbh, we need it to clean the Titan's clock and bring nVidia down a peg or 9

Unlikely as the TX packs 12gb of VRAM and will remain the best high resolution mGPU solution.

Having said that hardly anyone buys the TX, the real target for AMD is the GTX 980 Ti.

AMD need to do well with their next range of cards for everyones benefit or NVidia will turn into the intel of GPUs.