computerbaseAshes of the Singularity Beta1 DirectX 12 Benchmarks

Page 55 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Jaydip

Diamond Member
Mar 29, 2010
3,691
21
81
If you had pay attention to DX12, other games actually show regression on NV in DX12. Even in games they sponsor like Rise of Tomb Raider. In Ashes it's identical performance in DX11 vs DX12 here from Sweclockers (probably the 4x MSAA makes it very GPU bound at 1440p). Other sites find a small gain in DX12 for NV when Async Compute is disabled.

If you have problems with that, ask NVIDIA to enable Async Compute in their drivers. They still haven't and they promised it a long time ago last year.

You missed my point, even with async on Fury's lead is nothing amazing, AMD gains way more with DX 12 compared to NV and that is not attributed to async.Something is definitely broken here.
 

Mercennarius

Senior member
Oct 28, 2015
466
84
91
You missed my point, even with async on Fury's lead is nothing amazing, AMD gains way more with DX 12 compared to NV and that is not attributed to async.Something is definitely broken here.

Why must something be broken? They are different architectures so one is bound to respond differently than the other.
 

Jaydip

Diamond Member
Mar 29, 2010
3,691
21
81
Why must something be broken? They are different architectures so one is bound to respond differently than the other.

Including regression of performance? that's completely broken to me. I don't think MS designed dx 12 such that it will offer lower performance compared to dx 11, it makes no sense.
 

Mercennarius

Senior member
Oct 28, 2015
466
84
91
Including regression of performance? that's completely broken to me. I don't think MS designed dx 12 such that it will offer lower performance compared to dx 11, it makes no sense.

It doesn't have to have regressive performance. It has the potential to have significantly better performance as seen by AMD here on AOTS. DX12 evolved after Mantle with a similar approach, and Mantle was geared entirely around AMD's architecture. Microsoft also designed DX12 with the Xbox One in mind as it also supports DX12. And we all know that the Xbox One uses AMD's GCN architecture as well. So it's not that anything is broken but rather DX12 (And other recent APIs) are preferring to take advantage of the architecture AMD GPUs have taken and that means NVidia will have to play catch up on the next gen GPUs.
 

IllogicalGlory

Senior member
Mar 8, 2013
934
346
136
Every DX12 game so far shows regression compared to its DX11 path except when CPU limited or when async compute can be leveraged.
 
Feb 19, 2009
10,457
10
76
Including regression of performance? that's completely broken to me. I don't think MS designed dx 12 such that it will offer lower performance compared to dx 11, it makes no sense.

There is no regression, you can view Anandtech's article as it's more indepth. When they disable Async Compute for NV GPU, they gain some performance slightly.

Same for AMD GPUs. The gains get bigger when Async Compute is used, up to 20% for Fury X at 4K.
 

IllogicalGlory

Senior member
Mar 8, 2013
934
346
136
Correct! Every dx12 game has a performance regression compared to dx11, except when it doesn't...
Indeed. When GPU bound, Tomb Raider, Hitman and Ashes all display the same characteristics.

I don't understand the mockery; it's a perfectly valid response to Jaydip's assertion that DirectX12 should never result in a performance regression, yet we've seen that it does across multiple games, even NV-sponsored ones. i.e. we can conclude that that behavior is not a problem with Ashes's DX12 implementation, or at least nothing wrong with it that isn't wrong with every other DX12 game so far.

(not directed at you) It's interesting how many people will come forward to discredit and smear a studio that's designing a game with PC in mind, leveraging PC-only capabilities, and being wonderfully open just because their favorite company isn't winning. I guess what they want is more broken console ports with Gameworks taped on where you need a 980 Ti to get 45~60 FPS at 1080p.
 
Last edited:

Ma_Deuce

Member
Jun 19, 2015
175
0
0
just looking back a page to the hitman benches, I'm not seeing the regression. Are you just referring to nv gpus? I have seen regression a couple places, but nothing that strikes me as unusual for such a new api
 

IllogicalGlory

Senior member
Mar 8, 2013
934
346
136
Me neither, it was just a response to Jaydip. I think, in Hitman, AMD GPUs can leverage async to at least maintain parity with DX11 even in worst-case scenarios, but in those scenarios, its very close to DX11 performance, but the regression is fairly consistent with NV.
 
Feb 19, 2009
10,457
10
76
just looking back a page to the hitman benches, I'm not seeing the regression. Are you just referring to nv gpus? I have seen regression a couple places, but nothing that strikes me as unusual for such a new api

There is no regression. Not in Ashes for AMD DX11 vs DX12.

Not in Hitman.

Here's my own benchmark, R290X stock 1ghz.

DX11: AVG ~63 FPS (Min FPS occur at the start, scene load)



DX12: AVG ~80 FPS.



There's only regression in RoTR DX12 patch, but that's not ready as Nixxes has said.
 

Jaydip

Diamond Member
Mar 29, 2010
3,691
21
81
Me neither, it was just a response to Jaydip. I think, in Hitman, AMD GPUs can leverage async to at least maintain parity with DX11 even in worst-case scenarios, but in those scenarios, its very close to DX11 performance, but the regression is fairly consistent with NV.

If you look at the charts closely Async amounts to a minuscule performance improvement for Fury but going from dx 11 to dx 12 AMD sees a large performance jump while for NV it is virtually non existent. It can only happen if either NV's driver stack for dx 12 is not fully ready or the devs didn't optimize for NV that much.
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
If you look at the charts closely Async amounts to a minuscule performance improvement for Fury but going from dx 11 to dx 12 AMD sees a large performance jump while for NV it is virtually non existent. It can only happen if either NV's driver stack for dx 12 is not fully ready or the devs didn't optimize for NV that much.

Or their hardware just doesn't perform as well with DX12 relative to DX11. So, they don't get the performance boost.

It is possible that due to the consoles or Mantle -> DX12 similarities that AMD hardware is just better suited.
 

Jaydip

Diamond Member
Mar 29, 2010
3,691
21
81
Or their hardware just doesn't perform as well with DX12 relative to DX11. So, they don't get the performance boost.

It is possible that due to the consoles or Mantle -> DX12 similarities that AMD hardware is just better suited.

t is very much possible yes but on that other hand if the devs see regression of performance on NV hardware they should contact them as well.I am not saying that they didn't but from a api pov it makes no sense to have lower performance .
 

coercitiv

Diamond Member
Jan 24, 2014
6,400
12,852
136
Indeed. When GPU bound, Tomb Raider, Hitman and Ashes all display the same characteristics.

I don't understand the mockery; it's a perfectly valid response to Jaydip's assertion that DirectX12 should never result in a performance regression, yet we've seen that it does across multiple games, even NV-sponsored ones. i.e. we can conclude that that behavior is not a problem with Ashes's DX12 implementation, or at least nothing wrong with it that isn't wrong with every other DX12 game so far.
The way I understand it is that part of the way DX12 is built does lead to a synchronization (performance) cost, but this implementation allows for the use of other techniques (such as AC) in order to both compensate and bring overall performance increase. Maxwell is in a tough spot because it eats the performance cost but does not (yet?) benefit from any compensating feature of DX12. AC doesn't help, driver overhead was already very good in DX11 so there's no visible gain there either.

Saying that DX12 should never result in performance loss is a very broad statement, one we should make based on knowledge, not expectations. I for one, given my limited experience, would not venture to support it.
 

airfathaaaaa

Senior member
Feb 12, 2016
692
12
81
Including regression of performance? that's completely broken to me. I don't think MS designed dx 12 such that it will offer lower performance compared to dx 11, it makes no sense.
it doesnt make sense because nvidia never had any true hardware support for it despite all the lies they keep telling us
 

Bacon1

Diamond Member
Feb 14, 2016
3,430
1,018
91
What did they do to make this so personal?

Well considering they sold people hardware which stated support for it:



And then think its a "Fun Fact" when they don't actually support it:

Fun FACT of the day: Async Compute is NOT enabled on the driver-side with public Game Ready Drivers. You need app-side + driver-side!

https://twitter.com/PellyNV/status/702556025816125440

I'd be pretty pissed that I was mislead into buying hardware with false advertisement, and that they are laughing about it.

Not to mention they laughed about the whole 3.5/4 gb 970 issue instead of fixing their ads.



So the real question is, with Nvidia lying to your face about hardware features, Why aren't you mad at Nvidia?
 

BlitzWulf

Member
Mar 3, 2016
165
73
101
AOTS came out of early access today.Let the Benchmarking begin again!

INB4 someone says even though it's officially released that it doesnt represent "real" DX12 Games
 
Last edited:

BlitzWulf

Member
Mar 3, 2016
165
73
101
I think he said 390x?
He has a windows 7 system, so no direct x 12.

He did the 380 on win 10 earlier,honestly this guy is an awful streamer I'm kind of embarrassed to have linked it . I've been trying to watch for an hour and i think he's done 2 benches i'm not even sure most of the time he has been trying to set it up.He lost me when he switched to a windows 7 system.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |