• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD Polaris architecture – GCN 4.0

Its not about dies sizes, at all. In the current stack of GPU's, 980ti/FuryX is the high end... By the time you get £100 GPU's you are talking low end. The comparison AMD made was "console class" which is now sub-£100 GPU, how is that not low end, compared with a £500 product?

If highend is £500, mid range is £250-300, <£150 is low end... Why would we be expected to pay a higher price for a lower performing GPU, the term mid range is not compatible with "console class". A GTX 580 is no longer mid range or high end just because of its die size, that's nonsense.
 
Everyone has their own idea of mid-range - to me if it's over £250 it's high-end as it's more than I'd like to spend on a GPU. Low end is anything under about £125. (At 125 many users will feel they bought an expensive card)

Chip sizes on a process node have the advantage of being a way to compare that isn't totally subjective to the buyer.
 
Everyone has their own idea of mid-range - to me if it's over £250 it's high-end as it's more than I'd like to spend on a GPU. Low end is anything under about £125. (At 125 many users will feel they bought an expensive card)

Chip sizes on a process node have the advantage of being a way to compare that isn't totally subjective to the buyer.

Even if you remove prices from the equation, then the consumer would probably look at relative performance instead of die sizes. Its irrelevant to the end user what die size something is - but "console class" e.g. 1080p/30hz is low end, it is also well below £100 by this point.

Die size in and of itself is irrelevant to the buyer, they only care about what performance (upgrade) they get from a particular purchase... if the newer GPU is more expensive but lower performance then they aren't going to make that upgrade. There are APU's now that can do 720p/1080p/30hz, so how can anyone try to argue console class is midrange.
 
Last edited:
Its not about dies sizes, at all. In the current stack of GPU's, 980ti/FuryX is the high end... By the time you get £100 GPU's you are talking low end. The comparison AMD made was "console class" which is now sub-£100 GPU, how is that not low end, compared with a £500 product?

If highend is £500, mid range is £250-300, <£150 is low end... Why would we be expected to pay a higher price for a lower performing GPU, the term mid range is not compatible with "console class". A GTX 580 is no longer mid range or high end just because of its die size, that's nonsense.

Exactly. It's about performance and price.

dm seems to think that a mid-range card on release is a mid-range card throughout its lifetime. No. A card can start mid-range and be low-end a few years later. How this is even debatable baffles me.

Normally dm has some good points, but saying a 7870 is mid-range is today's market is just plain bonkers.

Today, a 7870 is low-end, and there is no argument.

Even if you remove prices from the equation, then the consumer would probably look at relative performance instead of die sizes. Its irrelevant to the end user what die size something is - but "console class" e.g. 1080p/30hz is low end, it is also well below £100 by this point.

Die size in and of itself is irrelevant to the buyer, they only care about what performance (upgrade) they get from a particular purchase... if the newer GPU is more expensive but lower performance then they aren't going to make that upgrade. There are APU's now that can do 720p/1080p/30hz, so how can anyone try to argue console class is midrange.

Totally agree. Console-class is not mid-range, it's low-end.
 
Last edited:
Even if you remove prices from the equation, then the consumer would probably look at relative performance instead of die sizes. Its irrelevant to the end user what die size something is - but "console class" e.g. 1080p/30hz is low end, it is also well below £100 by this point.

Die size in and of itself is irrelevant to the buyer, they only care about what performance (upgrade) they get from a particular purchase... if the newer GPU is more expensive but lower performance then they aren't going to make that upgrade. There are APU's now that can do 720p/1080p/30hz, so how can anyone try to argue console class is midrange.

Spot on. Only tech heads have an interest in the die size but even then I doubt that would make up their minds on purchasing decisions. Console class to me is low end also.
 
The other thing is, dm says 7870 is mid-range, well what about the 270? It's practically the same chip (some differences) and same performance.

Yet above the 270 at the time were the 280, 280x, 290, 290x. And today you also have the Fury above it also.

So is the 270 low-end or mid-range?

Well dm has his criteria but most here are going to judge based on performance, not die size, or how the card was placed at luanch (years ago). Because none of that stuff matters to us. What matters is how it performs vs the competition in today's market and today's lineup.

Therefore a 7870 performing card is not mid-range, but low-end.
 
A 270 is Mid range.

Low Range AMD
R7 240
R7 250
R7 260/X
R7 265

Low Range Nvidia
GT 210
GT 610
GT 720
GT 730

Mid Range AMD
R9 370/x
R9 380/x

Mid Range Nvidia
GTX 750/TI
GTX 650
GTX 660

High Range AMD
R9 390/X

High Range Nvidia
GTX 970
GTX 980

Enthusiast Range AMD
Fury Nano
Fury/X

Enthusiast Range Nvidia
GTX 980TI
GTX Titan-X
 
...comprises of two GPUs. A low-end, "console-class" GPU, and a mid-range that will match/slightly beat a Fury.

:p

This before the Summer, later in the year High and Enthusiast end GPU's from a more mature Samsung 14nm node or TSMC 16nm node.
 
Last edited:
This before the Summer, later in the Year High and Enthusiast end GPU's from a more mature Samsung 14nm node or TSMC 16nm node.

Only two GPUs in 2016. That's straight from AMD themselves.

Big Polaris is 2017 at the earliest.

Speculation is that the low-end, tiny Polaris (notebooks) is due in a couple months. With medium Polaris (desktop) about mid-2016. June/July.

Big Polaris is nowhere to be seen, and hasn't even taped out yet, apparently.
 
Only two GPUs in 2016. That's straight from AMD themselves.

Big Polaris is 2017 at the earliest.

Speculation is that the low-end, tiny Polaris (notebooks) is due in a couple months. With medium Polaris (desktop) about mid-2016. June/July.

Big Polaris is nowhere to be seen, and hasn't even taped out yet, apparently.

Did they say why? do you have a link? i hope this isn't true because if it is its also likely to be true for Nvidia.
 
http://fudzilla.com/news/graphics/39243-two-new-amd-gpus-in-2016

They are called Polaris 11 and Polaris 10 (see last page for links).

It doesn't say anything in that at all about 2017 releases or what size they are or other GPU's yet to tape out.

Just that there will be 2 GPU's in 2016, 2 GPU's could be two different architectures. like we currently have in GCN 1.1 and GCN 1.2 and they spread across Low to Enthusiast end.
One with and one without HBM?
 
Last edited:
Could happen easily. Both companies going the 970/980 then 980ti way. We'll get some low/mid range and mid/high range cards, and the bigger ones being a bit faster than the current top cards.
Then they bring the big guns later.
 
Could happen easily. Both companies going the 970/980 then 980ti way. We'll get some low/mid range and mid/high range cards, and the bigger ones being a bit faster than the current top cards.
Then they bring the big guns later.


Yeah, this is what i think is going to happen, on both sides.

Might keep my 970 for a bit longer, once the big guns are out the price of Fury-X / 980TI equivalents should drop.

I'm looking for Fur-X / 980TI < performance for £300, hopefully on the plus side of those.
 
Last edited:
Well obviously it will be 2 GPU's that will make up the entire line up of cards ( probably 3 GPU's with the largest core still to be shown/made at a latter date.)

It will actually be interesting as the previous AMD card line ups have had

7000 =7 separate GPU's making up 18 cards 6 of which are OEM only

200 =9 separate GPU's making up 20 cards 4 of which are OEM only.

300 =6 separate GPU's making up 16 cards 7 of which are OEM only.


We can only assume that the very very low end will be rebrands as is the norm.
 
Last edited:
...comprises of two GPUs. A low-end, "console-class" GPU, and a mid-range that will match/slightly beat a Fury.

:p

Please stop with the trolling and FUD. These two Polaris micro-architectures have been confirmed to cover everything from notebook to enthusiast level GPUs and other than some random comments from forumites (like you) there has been zero mention from AMD that a "big" Polaris is being help back until 2017.

http://wccftech.com/amd-unveils-polaris-11-10-gpu/
wccftech said:
Apart from this mobile-centric small Polaris GPU we were also made aware of an enthusiast version that I’d mentioned above which is also coming out. This “enthusiast” Polaris GPU has since been shown to journalists yesterday at CES. It has been described as the “successor” to the R9 Fury X and R9 390X graphics cards, so it’s clearly a high-end part. This is all part of AMD”s plan to release several SKUs based on each GPU to cover the entire market and regain market share Su affirmed. From entry level graphics products to mid-range and high-end parts.

Red bolded parts are contradicting the idea that they only plan a low end "console class" and a "mid-range". Stop spreading FUD that is purely your own subjective nonsense thrown around as a form of confirmation bias.
 
Last edited:
Back
Top Bottom