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

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

airfathaaaaa

Senior member
Feb 12, 2016
692
12
81
- Geforce GTX 1080
B2ZbzHn.jpg


- Geforce GTX 980 Ti @ 1525 MHz Core & 3950 Mhz Memory
9tMjWV2.jpg


Not bad considering Async is off for NVIDIA cards.



Let's all hope those Polaris 10 results are way off, because if GFXBench and AotS leaks reflect what to expect from this particular card (Ellesmere Pro?) I'm not very confident about it's performance. :(
it doesnt say anywhere that async is off tho
 

SPBHM

Diamond Member
Sep 12, 2012
5,068
422
126
that's far from impressive considering new process tech, new memory tech and high price.
 

Mopetar

Diamond Member
Jan 31, 2011
8,497
7,753
136
that's far from impressive considering new process tech, new memory tech and high price.

It's also only one game. Considering that async didn't work well on Maxwell I wouldn't be surprised if the developers didn't bother to do much optimization for NVidia cards.
 

airfathaaaaa

Senior member
Feb 12, 2016
692
12
81
It's also only one game. Considering that async didn't work well on Maxwell I wouldn't be surprised if the developers didn't bother to do much optimization for NVidia cards.
you know the whole "async" fiasco started because nvidia openly accused stardock for not developing the game for nvidia
then stardock released emails showing that they were talking with nvidia almost a year before the game was even a beta
and then all hell brake loose
 

tential

Diamond Member
May 13, 2008
7,348
642
121
It's also only one game. Considering that async didn't work well on Maxwell I wouldn't be surprised if the developers didn't bother to do much optimization for NVidia cards.
Yes they should have optimized for unreleased pascal by now Jesus tall about a terrible dev!
If only Nvidia had developer relations or like talked with the devs(which they did in this case for a long period of time).

Maxwell doesn't do well because maxwell isn't made for dx12 gaming. That's just plain and simple. It's no big deal but that's just the facts. It was a good product for its time, now dump it and move on.
 

Armsdealer

Member
May 10, 2016
181
9
36
There is a 5k/crazy benchmark in AotS in which the polaris part is 40% slower than titan X. That bench is definitely not cpu constrained since running on 5960x.

http://www.ashesofthesingularity.com/metaverse#/personas/8c77445b-f445-44ac-8dd5-401b55ffefeb/match-details/1e0a82e3-fcf3-4193-b25a-032cb80df081



What do you folks make of that? I'm inclined to believe this is likely not driver locked.
 

kondziowy

Senior member
Feb 19, 2016
212
188
116
Polaris 10 was demoed at GDC running Hitman DX12 1440p at 60fps exactly at this location: https://youtu.be/B_TJRSC7dOY?t=22. You can't do that with a chip slower than Fury X.

+Polaris 11 was demoed consuming 2x less energy than Maxwell. So why wouldn't 130W chip beat Fury X/ Titan X (250W). I think it will. You can call me crazy, but if it doesn't beat those, there is no reason to buy it.
 
Last edited:

digitaldurandal

Golden Member
Dec 3, 2009
1,828
0
76
Polaris 10 was demoed at GDC running Hitman DX12 1440p at 60fps exactly at this location: https://youtu.be/B_TJRSC7dOY?t=22. You can't do that with a chip slower than Fury X.

+Polaris 11 was demoed consuming 2x less energy than Maxwell. So why wouldn't 130W chip beat Fury X/ Titan X (250W). I think it will. You can call me crazy, but if it doesn't beat those, there is no reason to buy it.

Okay, so this keeps getting brought up, but as I recall didn't the representatives refuse to let anyone know the exact settings besides 1440p? Was it running on Medium? High? Ultra? I don't have the game or I would be tempted to compare screenshots to different settings.

Also was it running "at 60 fps" or running at "never below 60 fps" because those are different things.

Also don't we think that if AMD had such an incredibly performing card for that size they would be promoting that, especially after Nvidia's sidegrade announcement? We already know that Fury is going to exist in the market with Polaris, I'm really not sure how people keep thinking Polaris is going to replace it.
 

Keysplayr

Elite Member
Jan 16, 2003
21,219
55
91
There is a 5k/crazy benchmark in AotS in which the polaris part is 40% slower than titan X. That bench is definitely not cpu constrained since running on 5960x.

http://www.ashesofthesingularity.co...-details/1e0a82e3-fcf3-4193-b25a-032cb80df081



What do you folks make of that? I'm inclined to believe this is likely not driver locked.

If all it took to make any card play anything at any res was to lock the driver, then all the vendors would have done it. Just locking a driver at 40fps doesn't guarantee you'll get 40 fps at ANY resolution. Like Vsync, your fps will stay at the vsync setting until the card can't cope. Then inevitably, the fps drastically drops a tier.
 

xthetenth

Golden Member
Oct 14, 2014
1,800
529
106
There is a 5k/crazy benchmark in AotS in which the polaris part is 40% slower than titan X. That bench is definitely not cpu constrained since running on 5960x.

http://www.ashesofthesingularity.co...-details/1e0a82e3-fcf3-4193-b25a-032cb80df081



What do you folks make of that? I'm inclined to believe this is likely not driver locked.

Is that the one where it's in a set of results getting the same framerate at a variety of resolutions?

The real problem is that it's very easy to get hardware, especially prerelease hardware, to underperform, and a run where it's underperforming is basically irrelevant. It may just be that it falls off a cliff at 5k, but it performs great up until that point. It may be running underclocked. It may be running drivers that severely underutilize it. It could be any, all or none of those things and we just can't know as of now. Unfortunately you can't establish subpar performance for sure without proper release benchmarks where you have confidence about the setup, you can only establish good performance with a bit of fudge factor to account for the possibility of a top percentile overclock (since if it can do that performance you know it can do at least that performance).

I've got to lay off the ~month before launch speculation and rumors, all the grains of salt are going to give me a coronary. :D
 

Layer8

Member
May 3, 2016
43
0
6
What do you folks make of that?
Depends on which benchmark run you are looking at.

It is 39% slower than that one Titan X in one of the benchmarks.
But it also is just 41% slower than those 2 Titan X in the other benchmark.

So what do I make of it?
There is just not enough information to really compare those benchmarks.
Pick the one you like for comparison.
As usually, the truth is most likely somewhere in between.
 

hawtdawg

Golden Member
Jun 4, 2005
1,223
7
81
This is not even a tock, its just Maxwell adjusted for maximum performance/area @ 16nm, given TSMC process characteristics.

Take GTX980, add 512 SP and clock it @ 1800Mhz, change GDDR5 by GDDR5X and bingo :thumbsup:

1070 literally just looks like a 980 with higher clocks.
 

airfathaaaaa

Senior member
Feb 12, 2016
692
12
81
well you can literally see 1080s from 27 average fps till 50 fps i imagine that 50 fps must be a highly clocked one
 

AtenRa

Lifer
Feb 2, 2009
14,003
3,362
136
Just for the data, I get 38fps on 1080p Crazy with Core i7 3770K@4.44GHz and R9 390 default.
 

AtenRa

Lifer
Feb 2, 2009
14,003
3,362
136
What about 5K crazy, if you don't mind?

I can do up to 3200x1800 with VSR

Core i7 3770K@4.44GHz
R9 390 Cat 16.5.1
Win 10 Pro 64bit

1080p Crazy 38fps average
2560x1440 Crazy 34fps average
3200x1800 Crazy 29fps average
 

boozzer

Golden Member
Jan 12, 2012
1,549
18
81
Dunno what happened to the official threads, but using this thread since same source website.

Seems AMD is maybe pushing Vega 10 to October release? Hmmmm...interesting.

http://videocardz.com/59808/amd-vega-gpu-allegedly-pushed-forward-to-october

Raining on NV's parade or they didn't think GP104 would be that powerful to be left uncontested for so long?

The plot thickens!!!
I am having conflicting feelings about this. If polaris 10 n 11 sucks, that means my 290 will be gold for another year because 1070 would also suck. But I get to save monies. kinda weird feelings to be having at the same time.
 

railven

Diamond Member
Mar 25, 2010
6,604
561
126
I am having conflicting feelings about this. If polaris 10 n 11 sucks, that means my 290 will be gold for another year because 1070 would also suck. But I get to save monies. kinda weird feelings to be having at the same time.

Save money where you can. It's the golden rule! I cut out my second cup of coffee (work coffee sucks anyways, garbage) and started leaving a thermo (from home's morning coffee) in the car. It's nice and cool, taste better, and considering my caffeine addiction, I've saved a few hundreds since I started this new endeavor.

Nickle here and there.