AMD Polaris architecture – GCN 4.0

[Seanspeed]

A joking sort of calculation? he was not joking about 10x Maxwell.

Yes, he absolutely was. He made abundantly clear for anybody paying attention when he said this number came by using 'CEO math' and they explained exactly where the number came from and it didn't have anything to do with a realistic 10x improvement in gaming performance.

I'm quite sure that somebody who is as clued-in as you are knows all this though, and are just playing dumb in order to have a go.

And it still doesn't make your comment the least bit relevant to what I said. It was just a 'return attack' at what you obviously perceived as an attack on AMD, which was not my intention at all.

 

[Orangey]

I told you LPP was garbage low power node unsuitable for GPUs.

Yet more proof they aren't using Samsung too (in addition to the fact they themselves said "only GF"). I'm sure it's Global's fault at least partially.

For those that don't know, AMD have a wafer supply agreement with this foundry. They get fined for not buying a certain amount. Why don't they get out of it? The people who own the foundry sit on AMD's board of directors. That is why AMD keep using these clowns year after year. Seems AMD at least lucked by NV having their own woes with finfets. So no monopoly just yet.

 

[humbug]

Yes, he absolutely was. He made abundantly clear for anybody paying attention when he said this number came by using 'CEO math' and they explained exactly where the number came from and it didn't have anything to do with a realistic 10x improvement in gaming performance.

I'm quite sure that somebody who is as clued-in as you are knows all this though, and are just playing dumb in order to have a go.

And it still doesn't make your comment the least bit relevant to what I said. It was just a 'return attack' at what you obviously perceived as an attack on AMD, which was not my intention at all.

Are you serious or is that a bad attempt at a joke?

Genuine question. Explain yourself.

 

[LoadsaMoney]

If they do come end of this month, then Nvidias marketshare will start evaporating, so they'll need to get their lower Pascal cards out quick, as AMD will just be killing em

 

[Jono8]

Are you serious or is that a bad attempt at a joke?

Genuine question.

Did Nvidia explicitly state it was 10x faster in terms of relative gaming performance?

 

[Seanspeed]

Are you serious or is that a bad attempt at a joke?

Genuine question. Explain yourself.

I've explained all that needed explaining. :/

He was completely open about it not being a serious calculation(hence the 'CEO math' comment) and went on to say exactly how the number came to be, which should have made it obvious to anyone with an inkling of how this works that it didn't mean 10x gaming performance or anything of the sort.

Is this how the discussion is going to go here? You play dumb and you make me write out a bunch of stuff you already damn well know just so you can continue to ignore my initial comment and distract with this lousy attempt at deflection?

 

[humbug]

I'm not playing dumb ^^^ i'm wondering if you're playing dumb, or not?

Did Nvidia explicitly state it was 10x faster in terms of relative gaming performance?

Good Grief... Where did i say he did?

Chill out guys, there just GPU's.

 

[JediFragger]

If they do come end of this month, then Nvidias marketshare will start evaporating, so they'll need to get their lower Pascal cards out quick, as AMD will just be killing em

£200 and 980Ti performance and AMD will indeed 'kill it'

Knowing AMD's current mindset though it'll more than likely be £275+

 

[andybird123]

Do you have a link to that?

I this one they list 50% reduction in size http://www.samsung.com/semiconductor/foundry/process-technology/14nm/
are they saying different things at different times or do they have different links with differing info?

The iriginal press release said 0.55 tines the die area (1.9 times the density) cant find the original samsung release but if you google lpp 0.55 you will find all the tech sites reporting it

Even 50% is not 2.3 times, its a straight 2x

 

[Seanspeed]

I'm not playing dumb ^^^ i'm wondering if you're playing dumb, or not?

"I know you are but what am I?"

Is that the level of discourse I should expect from you when I challenge your opinion?

Continue deflecting man. This is really sad.

 

[D.P.]

Directly from AMD that claimed during their live event this year that Polaris will offer up to 2.5X the performance per watt of their current line up.

That is not transistor density though.

And fi you want to make prediction based on IHV performance-per-watt changes then nvidia have claimed 3x performance per watt for Pascal. Read the fine-print and grab a big bucket fo salt with any of these changes.

 

[Jono8]

I'm not playing dumb ^^^ i'm wondering if you're playing dumb, or not?



Good Grief... Where did i say he did?

Chill out guys, there just GPU's.

Say what?

I'm already disappointed, not so long ago Pascal was 10x Maxwell, Jens 'the man himself' said so, then suddenly its only 2x Maxwell, and then we learn its that in a very very rough sense under special circumstance with a lot of things exaggerated.

After all that it now looks like what we actually have is Maxwell x1.25.... from 10x to that, what an anticlimax. so disappointed. eh?

 

[humbug]

Right... i said nothing about Gaming Performance ^^^^^

Did Nvidia explicitly state it was 10x faster in terms of relative gaming performance?

"I know you are but what am I?"

Is that the level of discourse I should expect from you when I challenge your opinion?

Continue deflecting man. This is really sad.

You did not challenge my opinion, you accused me of perceiving AMD as under attack and reacting to that.

A completely bizarre and insane reaction on your part not only because of whats in my signature but also nothing to do with anything i have said, or at the very least completely miss represented by you.

Its not just off topic and personal on your part, its completely mad.

The problem here is Jens is not my God. and that can't be tolerated. its like religious extremes in here. everyone just chill out.

 

[Seanspeed]

You did not challenge my opinion, you accused me of perceiving AMD as under attack and reacting to that.

No man, my initial comment was about you playing up Polaris 10 expectations and acting completely assured of it. I have no problem with somebody thinking something *might* be at whatever level, but you seem to completely dismiss any idea that it might not be as good as you're thinking, either.

That's where our whole conversation started and that's where I challenged your opinion. Then you went on about some nonsense with something the Nvidia guy said in response and you've successfully deflected the topic since then.

The problem here is Jens is not my God. and that can't be tolerated.

So I was 100% correct in you taking my comment as an attack on AMD as you've pigeonholed me as some Nvidia fanboy and obviously that could be my only reason for posting what I did. Even though I actually would LOVE for you to be right as I've got no bone in any fight and would gladly get a P10 card if it's as good as you're saying it absolutely will be.

 

[Jono8]

Right... i said nothing about Gaming Performance ^^^^^



.

Yes you did, you directly compared that 10x statement to what we know about the relative gaming performance of the 1080 ( ie 1.25 against a 980Ti/Titan X).

Otherwise your whole sentence below was nonsense.

"After all that it now looks like what we actually have is Maxwell x1.25.... from 10x to that, what an anticlimax. so disappointed"

 

[faceman123]

I wonder how long AMD will survive, Nvidia have established themselves as the far superior competitor, performance aside, The nvidia cards feel built far better and more often then not hold their values better. Even if AMD were to bring out a superior card it wouldn't have the same impact as nvidia does. People have lost trust in the AMD brand. Look at the Figures, numbers don't lie.

 

[Mauller]

Yes you did, you directly compared that 10x statement to what we know about the gaming performance of the 1080 ( ie 1.25 against a 980Ti/Titan X).

Otherwise that whole post was nonsense.

"After all that it now looks like what we actually have is Maxwell x1.25.... from 10x to that, what an anticlimax. so disappointed"

I think one too many people are not understanding the satire of that statement.

 

[humbug]

Can you not see how obviously pedantic this is on my part, really? does it really need a disclaimer citing my sarcasm and thus should be taken as a playful half truth teas?

Jono8 and Seanspeed, get a grip.

I'm already disappointed, not so long ago Pascal was 10x Maxwell, Jens 'the man himself' said so, then suddenly its only 2x Maxwell, and then we learn its that in a very very rough sense under special circumstance with a lot of things exaggerated.

After all that it now looks like what we actually have is Maxwell x1.25.... from 10x to that, what an anticlimax. so disappointed. eh?

moving on....

 

[Lokken86]

I wonder how long AMD will survive, Nvidia have established themselves as the far superior competitor, performance aside, The nvidia cards feel built far better and more often then not hold their values better. Even if AMD were to bring out a superior card it wouldn't have the same impact as nvidia does. People have lost trust in the AMD brand. Look at the Figures, numbers don't lie.

Not much difference in build quality thread days tbh.

 

[Jono8]

I think one too many people are not understanding the satire of that statement.

Haha, what?

He can't have it both ways. He cant at one point act like he knew the 10x remark by Nvidia was disingenuous and then later down the line, argue when someone else says it was indeed disingenuous.