videocardzGTX 1080, Polaris 10/11 Ashes of Singularity DirectX12 benchmarks leaked

[Fallen Kell]

Originally Posted by Noctifer616
So it's like testing the GPU without working drivers? I guess that could explain this odd behavior, also kinda makes the testing invalid?

Or it even is intentionally gimped in drivers to lead to confusing leaks...

Seriously, this is just normal operating procedure. I mean, really, when was the last time AMD had working drivers within the first 6 months of a new card release? (That was rhetorical and a trick question, as the answer to it is "never", the last time was when it the company was still ATI, and thus not AMD).

 

[geoxile]

If we're putting in guesses I'd guess Polaris 10 will be 20-50% better than 390X based on the supposed 130W TDP and 2.5x improvement over the last generation. The guess depending on what the 2.5x improvement is compared to: 390X's Grenada or Fiji (in the Fury Nano). Even account for HBM the Fiji found in the Fury Nano should have impressive efficiency over Grenada, which is just Hawaii with a new name.

There's also the Primitive Discard Accelerator to consider. And improved shader efficiency, which may be power efficiency or hardware usage. If Pascal's frequency improvements are a result of the process, I expect Polaris to have similar, or slightly worse, frequency improvements but probably higher relative base clocks than GP104 prodducts.

 

[swilli89]

If we're putting in guesses I'd guess Polaris 10 will be 20-50% better than 390X based on the supposed 130W TDP

I hope this is true but from what most have said, 150 watts is going to be the upper end based on the 6-pin configuration, it remains to be seen how close AMD would want to come to that limit.

Whats the closest we've ever come to the 6-pin power limit. Has AMD/NV ever released a 6-pin card with a TDP of 150w? Really curious about that one.

There's also the Primitive Discard Accelerator to consider. And improved shader efficiency, which may be power efficiency or hardware usage. If Pascal's frequency improvements are a result of the process, I expect Polaris to have similar, or slightly worse, frequency improvements but probably higher relative base clocks than GP104 prodducts.

I think based on what I've read about the processes we can expect GloFo/Samsung to have a ~10% density advantage.

Apple spares no expenses designing its chips, after all; they are some of the highest grossing chips in human history! I have no doubt Apple had two dedicated teams for the two fabs and their respective designs for A9. Here we see Samsung with a 9.2% density advantage.

Copied from Chipworks

TSMC 16FF vs Samsung 14LPE


From Tomshardware

http://www.tomshardware.com/news/iphone-6s-a9-samsung-vs-tsmc,30306.html

Based on the results of our testing, it's clear that both versions of Apple's A9 SoC deliver the same level of performance, but Samsung's 14nm FinFET process appears to offer slightly better power efficiency, extending battery life between 3.5-10.8 percent.

So according to them, Samsung's process has better leakage, which would make me think they can clock at least as well as TSMC. However, this can't be assumed too much as Apple most likely designed for lower clockspeeds as they have always done.

Remember, this is all on 14LPE. We know AMD is using the Gen-II version of Samsung's FinFet 14LPP which should improve clockspeed/leakage even further.

 

[JDG1980]

I hope this is true but from what most have said, 150 watts is going to be the upper end based on the 6-pin configuration, it remains to be seen how close AMD would want to come to that limit.

Whats the closest we've ever come to the 6-pin power limit. Has AMD/NV ever released a 6-pin card with a TDP of 150w? Really curious about that one.

The Pitcairn-based R9 270 (non-X) had a TDP of 150W and a single 6-pin power connector. On the professional side, the Pitcairn-based FirePro W7000 and, later, the Tonga-based FirePro W7100 also had a 150W TDP and one 6-pin.

It would be surprising if AMD didn't go up to 150W on at least one Polaris 10 SKU. They may go higher - remember, Pitcairn-based R9 270X went all the way up to 180W. The only way this won't happen if is they run into severe diminshing returns (memory bottlenecks?) at much lower core clocks.

 

[Armsdealer]

Gentlemen,

Take a look at the Crazy 5k ranking on AotS which I posted earlier:

http://www.ashesofthesingularity.com/metaverse#/ladders/benchmark/overall/Crazy_5K?viewType=myself

Take note of the Polaris entrant run by npi3d_ashes and the time the benchmark was run (5/9/16 10:25:52)

Then take a look at the Crazy 4k ranking on AotS:

http://www.ashesofthesingularity.com/metaverse#/ladders/benchmark/overall/Crazy_4k?viewType=myself

Scroll down to #9 (npi3d_ashes). This is the same fellow who tested Polaris in 5k. Take a look at his 4k benchmark on what is ostensibly a dual Fury setup. Take note of the time the bench was run (5/9/16 10:25:52).

Click his profile to see his other benchmarks. Select the one run on 5/9/16 10:25:52. It shows a Radeon HD 5570 on a Core Duo that gets 11 fps.

What in the world is going on here?

 

[airfathaaaaa]

nothing more than hide and seek

 

[Armsdealer]

nothing more than hide and seek

In the sense that they've spoken with the AotS website trying to hide the Polaris benchmarks, right?

 

[thesmokingman]

Gentlemen,

Take a look at the Crazy 5k ranking on AotS which I posted earlier:

http://www.ashesofthesingularity.com/metaverse#/ladders/benchmark/overall/Crazy_5K?viewType=myself

Take note of the Polaris entrant run by npi3d_ashes and the time the benchmark was run (5/9/16 10:25:52)

Then take a look at the Crazy 4k ranking on AotS:

http://www.ashesofthesingularity.com/metaverse#/ladders/benchmark/overall/Crazy_4k?viewType=myself

Scroll down to #9 (npi3d_ashes). This is the same fellow who tested Polaris in 5k. Take a look at his 4k benchmark on what is ostensibly a dual Fury setup. Take note of the time the bench was run (5/9/16 10:25:52).

Click his profile to see his other benchmarks. Select the one run on 5/9/16 10:25:52. It shows a Radeon HD 5570 on a Core Duo that gets 11 fps.

What in the world is going on here?

I've been saying this all along. Rumor threads created based upon random forum posts as the true source is bad.

 

[airfathaaaaa]

In the sense that they've spoken with the AotS website trying to hide the Polaris benchmarks, right?

why? they could very well being testing clocks to hide it

 

[tential]

I've been saying this all along. Rumor threads created based upon random forum posts as the true source is bad.

I can't believe these sites actually use forum posts as sources.

 

[Armsdealer]

I can't believe these sites actually use forum posts as sources.

I actually don't know what conclusion you are drawing from my post because I don't know what conclusion *I'm* drawing from my post..

I'm just saying that there's some unusual behavior on the website surrounding the Polaris benchmarks. This is factual and can be proven from inspection of the site. Perhaps someone here understand why.

 

[Despoiler]

I can't believe these sites actually use forum posts as sources.

Gotta rake in that advertising money.

 

[thesmokingman]

I actually don't know what conclusion you are drawing from my post because I don't know what conclusion *I'm* drawing from my post..

I'm just saying that there's some unusual behavior on the website surrounding the Polaris benchmarks. This is factual and can be proven from inspection of the site. Perhaps someone here understand why.

He's concurring with you, us... its not kosher the way tech news draws upon unconfirmed random forum posters. It's like ouroboros...

 

[Armsdealer]

He's concurring with you, us... its not kosher the way tech news draws upon unconfirmed random forum posters. It's like ouroboros...

Got it. Either way, I sincerely hope GloFo didn't screw over AMD with their new process. It is genuinely bizarre that other than the Hitman demo we have no evidence EVER of GloFo's new node producing anything the client (eg Apple) was happy with. AMD has gone all-in with these buggers between zen, Polaris, and vega... I hope they did their homework and checked it twice.

 

[airfathaaaaa]

I actually don't know what conclusion you are drawing from my post because I don't know what conclusion *I'm* drawing from my post..

I'm just saying that there's some unusual behavior on the website surrounding the Polaris benchmarks. This is factual and can be proven from inspection of the site. Perhaps someone here understand why.

unusual as in 3 yellow jurnalistic articles in a matter of days concerning polaris? and all based on forum posts? well they are either blatanly bias so much that they dont even care anymore
or they are doing mistakes...
and given the history of guru3d and videocardz we all know what they are

 

[boozzer]

Seriously, this is just normal operating procedure. I mean, really, when was the last time AMD had working drivers within the first 6 months of a new card release? (That was rhetorical and a trick question, as the answer to it is "never", the last time was when it the company was still ATI, and thus not AMD).

that is why dx12 or any of the new api base on amd's api is a godsend. no longer will gamers depend on either nv or amd for drivers every damn time a game is release. :thumbsup:

 

[Abwx]

Got it. Either way, I sincerely hope GloFo didn't screw over AMD with their new process. It is genuinely bizarre that other than the Hitman demo we have no evidence EVER of GloFo's new node producing anything the client (eg Apple) was happy with. AMD has gone all-in with these buggers between zen, Polaris, and vega... I hope they did their homework and checked it twice.

If GF screwed something then there s tons of money to be made out of sueing them in courts, here is the prove that they eventualy made fraudulous claims to lure unsuspecting customers :

 

[Flapdrol1337]

that is why dx12 or any of the new api base on amd's api is a godsend. no longer will gamers depend on either nv or amd for drivers every damn time a game is release. :thumbsup:

Dunno, the 285 and fury were slower under mantle than dx11, some work still has to be done, either on the thin dx12 driver or by the game developer.

 

[Sunaiac]

when was the last time AMD had working drivers within the first 6 months of a new card release

Always.

I bought a 7970 that had working drivers. It was 30% faster than a same priced 580, and is now 30% faster than a much more expensive 680.
I bought a 290X that had working drivers. It was as fast as a 50% more expensive 780Ti, and is now as fast as a much more expensive and younger 980.

Trying to point out that AMD drivers are not as good as they could be when the cards come is a self defeating argument only making the cards look even more interesting, since the price/perf ratio you buy them is based on those performances.

 

[flopper]

Always.

I bought a 7970 that had working drivers. It was 30% faster than a same priced 580, and is now 30% faster than a much more expensive 680.
I bought a 290X that had working drivers. It was as fast as a 50% more expensive 780Ti, and is now as fast as a much more expensive and younger 980.

Trying to point out that AMD drivers are not as good as they could be when the cards come is a self defeating argument only making the cards look even more interesting, since the price/perf ratio you buy them is based on those performances.

Yea, to many noobs around here.
I guess the Nvidia 580 owners that bought cards that couldnt use sli for over a year with win vista blamed Microsoft than the guys from nvidia?

 

[airfathaaaaa]

Seriously, this is just normal operating procedure. I mean, really, when was the last time AMD had working drivers within the first 6 months of a new card release? (That was rhetorical and a trick question, as the answer to it is "never", the last time was when it the company was still ATI, and thus not AMD).

who has a driver that uses the full potential of the card right away? no one...not a single card came out using its full potential EVER

it took amd 2 years to bring 4850 to +50% from what it was
and look at all the gcn cards...5 years and still going and going..

 

[Spanners]

Seriously, this is just normal operating procedure. I mean, really, when was the last time AMD had working drivers within the first 6 months of a new card release? (That was rhetorical and a trick question, as the answer to it is "never", the last time was when it the company was still ATI, and thus not AMD).

Never had any issues myself, certainly nothing that would rise to level of their drivers not "working". Whatever that's intended to mean. Rhetorical and hyperbolic but I'm not seeing the trick? :thumbsdown:

 

[96Firebird]

who has a driver that uses the full potential of the card right away? no one...not a single card came out using its full potential EVER

it took amd 2 years to bring 4850 to +50% from what it was
and look at all the gcn cards...5 years and still going and going..

Whaaaattttt???

I've been told it is not AMD gaining performance via drivers, but Nvidia neglecting their cards to make the new ones look better?!?! Which one is it?

Sorry about the pointless underlining/bolding of random words, seems like the thing to do here nowadays.

 

[airfathaaaaa]

Whaaaattttt???

I've been told it is not AMD gaining performance via drivers, but Nvidia neglecting their cards to make the new ones look better?!?! Which one is it?

Sorry about the pointless underlining/bolding of random words, seems like the thing to do here nowadays.

nvidia can gimp their superscalar cards on dx11 and you know how we know that?
look how kepler cards are doing under aots...a freaking 780ti till 2k is beating 970 and its almost on par with 980...

 

[96Firebird]

A card with better compute is doing better in a game that uses compute?! Say it ain't so!

Do I need to demand proof from you? I know how much you love proof, but I don't want to embarrass you...