AMD's next GPU uarch is called "Polaris"

[Techhog]

All aboard the hype train!

Have we learned nothing from the past...

Well, it matched their demonstration unless they manipulated the results...

 

[Seba]

http://www.anandtech.com/show/9886/a...u-architecture
Quote:
Originally Posted by anandtech
To that end RTG also plugged each system into a power meter to measure the total system power at the wall. In the live press demonstration we saw the Polaris system average 88.1W while the GTX 950 system averaged 150W. Meanwhile in RTG’s own official lab tests (and used in the slide above) they measured 86W and 140W respectively.

Average. Meaning average between idle and full load? Then could be true, but also meaningless.

 

[Glo.]

Well, if this is true it means that GPU with performance levels at 1080p between R9 285 and 280X will use 30W.

WTF?

Average. Meaning average between idle and full load? Then could be true, but also meaningless.

I think what they meant was the average at gaming load.

 

[Seba]

Then why would they make a desktop graphics card with a ~30W power draw?

 

[Techhog]

Well, if this is true it means that GPU with performance levels at 1080p between R9 285 and 280X will use 30W.

WTF?


I think what they meant was the average at gaming load.

No, it would be more like R9 270 performance, and likely a TDP of 45W.

 

[Glo.]

Then why would they make a desktop graphics card with a ~30W power draw?

Apple.

 

[Glo.]

No, it would be more like R9 270 performance, and likely a TDP of 45W.

You are right, I've checked the OC'ed versions of GTX950.

 

[Kenmitch]

All aboard the hype train!

Have we learned nothing from the past...

It's not really possible to predict the future based on the past.

Not saying the hype train won't derail....Just saying we won't know till it does or doesnt.

 

[Mondozei]

The primitive discard accelerator is the real deal here. It can give a huge minimum fps boost, especially on complex scenes.

Please elaborate.

 

[Jakobx]

Well, if this is true it means that GPU with performance levels at 1080p between R9 285 and 280X will use 30W.

WTF?

I think what they meant was the average at gaming load.

err no...They compared it to 950 which is nowhere near 280x. 960 is the chip that is close to 280x / r380.

 

[gamervivek]

People take marketing slides as truth now. Seriously?

The slide is backed up by the live press video.

Wow.

https://www.youtube.com/watch?v=5g3eQejGJ_A&feature=youtu.be

AT article,

"In the live press demonstration we saw the Polaris system average 88.1W while the GTX 950 system averaged 150W"

 

[ShintaiDK]

People already forgot the Fiji "scores" they manipulated? Dont put much into it. Because its certain that reviews will show something different. Marketing never matches reality.

 

[DarkKnightDude]

Wait, so its 14nm? Not 16nm? Then that earlier rumor was way wrong.

 

[sontin]

"Wow"?
Maybe i miss something: AMD is comparing a 16nm FinFet chip to a 12months old chip with nearly 2x times the transistors in a AMD biased game. They archived a 2,5x improvement in efficiency.

nVidia released the GTX980 11 months after the 290X, which has less transistors and archived a 1.9x to 2.0x improvement...
Would have nVidia demonstraded the GTX980 in march 2014 nobody would believed it.

Yet when AMD needs a much longer timeframe, a new process node and nearly twice the transistors, everybody is screaming "Wow".

I guess the hype train is back and now bigger and better than ever.

 

[Head1985]

Also, after those new slides were published, it seams one of the new dies will be very small. Something like 100mm2 ??

So, what about the second die ??? will it be close to 200mm2 or even higher ???

They need GPU around 300-350mm2 to counter midrange pascal GTX970/980 successor.
If AMD add 80% more transistors that mean 20% less size vs hawaii.
That GPU should be 350mm2 and have 80% more performance vs 390x.

Same witch pascal
If Nv add 80% more transistors and lower size by 20% that mean that GPU will be 320mm2 with again 80% more performance vs GTX980.

 

[Noctifer616]

Wait, so its 14nm? Not 16nm? Then that earlier rumor was way wrong.

Both actually according to the AT article. Some GPU's will be made at GF 14nm and others are TSCM 16nm.

 

[Kenmitch]

"Wow"?
Maybe i miss something: AMD is comparing a 16nm FinFet chip to a 12months old chip with nearly 2x times the transistors in a AMD biased game. They archived a 2,5x improvement in efficiency.

nVidia released the GTX980 11 months after the 290X, which has less transistors and archived a 1.9x to 2.0x improvement...
Would have nVidia demonstraded the GTX980 in march 2014 nobody would believed it.

Yet when AMD needs a much longer timeframe, a new process node and nearly twice the transistors, everybody is screaming "Wow".

I guess the hype tran is back and now bigger and better than ever.

It's 14nm FinFet.

It's a demo with early working silicone on a bottom of the line GPU. There really isn't anything to get upset about....Yet....Or maybe at all.

It's in the best interest of all if AMD does bring on some serious heat this go around....Cheaper prices on your preferred NVidia products in the end.

 

[Trumpstyle]

So around 3x improved power efficiency I'm guessing from reading the anandtech articel. The new Gpu drew around 45W and had the perfomance of geforce 950/radeon 270x. It might even be more than 3x improved power efficiency. This seems atleast better than we all expected.

 

[railven]

All aboard the hype train!

Have we learned nothing from the past...

Being late to this thread, and after reading the AT article, I was slowly buying into the hype (someone here said 390/970 @ 86W, but I recalled AT article said it was a GTX 950).

Oh well, I'm still going to get on. Nothing else to do around here

Choo choo!!!

 

[gamervivek]

"Wow"?
Maybe i miss something: AMD is comparing a 16nm FinFet chip to a 12months old chip with nearly 2x times the transistors in a AMD biased game. They archived a 2,5x improvement in efficiency.

nVidia released the GTX980 11 months after the 290X, which has less transistors and archived a 1.9x to 2.0x improvement...
Would have nVidia demonstraded the GTX980 in march 2014 nobody would believed it.

Yet when AMD needs a much longer timeframe, a new process node and nearly twice the transistors, everybody is screaming "Wow".

I guess the hype train is back and now bigger and better than ever.

Yes, you missed the fact that the 'wow' was for AMD showing off working silicon to press and not just slides.

Being late to this thread, and after reading the AT article, I was slowly buying into the hype (someone here said 390/970 @ 86W, but I recalled AT article said it was a GTX 950).

Oh well, I'm still going to get on. Nothing else to do around here

Choo choo!!!

86W for total system power consumption. At the wall.

Hype train still on tracks, choo choo!!

 

[sontin]

So, showing of working silicon for a product released in 6 months is impressive?

nVidia had never shown Maxwell prior the launch and yet Maxwell was impressive at launch.

 

[Kenmitch]

Being late to this thread, and after reading the AT article, I was slowly buying into the hype (someone here said 390/970 @ 86W, but I recalled AT article said it was a GTX 950).

Oh well, I'm still going to get on. Nothing else to do around here

Choo choo!!!

That was total system power not just the cards. Whatever the little chip is/was it was only drawing around 40w or so....There is still plenty of watts for left for the hype train.

So, showing of working silicon for a product released in 6 months is impressive?

nVidia had never shown Maxwell prior the launch and yet Maxwell was impressive at launch.

You would rather see slides? For me the showing of the most likely bottom of the card stack as early working silicone is much better.

 

[railven]

86W for total system power consumption. At the wall.

Hype train still on tracks, choo choo!!

I know, and that's what made that prior post even more amazing to me!

This next node is desperately needed. Come mid-2016 my card will be a year old and rusty.

 

[ShintaiDK]

Wait, so its 14nm? Not 16nm? Then that earlier rumor was way wrong.

AMD 14LPP, NVidia 16FF+.

 

[gamervivek]

So, showing of working silicon for a product released in 6 months is impressive?

nVidia had never shown Maxwell prior the launch and yet Maxwell was impressive at launch.

Yes. Not for your of course.

nvidia missed the hype train, too bad.