First Nvidia & AMD Deus Ex: Mankind Divided DX12 Benchmark appeared

[i_build_nanosuits]

Just now, Pohernori said:

 

Video shows only the 1060 and 480. 

 

Irrelevant to what I am concerned about. 

what are your concern?!

980ti vs Fury X? same deal...both loose around 10 FPS from going to DX12.

[Notional]

19 hours ago, That Norwegian Guy said:

not actually a bug, it's a feature of the Nvidia drivers that forces Pascal cards out of concurrent async compute if the game tries to run it (because they can't) into context switching

Yeah once again NVidia is outed by how poor their DX12/Vulkan support really is. For AMD to move the entire playing field with low level API's was brilliant. 

2 hours ago, i_build_nanosuits said:

Yes they did...it's an AMD game.

Also, it's bullshit...Joker tested it with the latest AMD driver and the radeon card lost nearly 10 FPS by going with DX12 and so does the nvidia card, here:

 

Joker must have derped.

1 hour ago, i_build_nanosuits said:

AMD loose less performance by going to DX12...way to go. lol

I gain 10-15% FPS across the field from min to max fps.

20 hours ago, Hakugman said:

The whole DX 12 API is broken as hell atm, don't get your panties in a bunch. I could say that Microsoft should learn from Vulkan and get their game together because so far their marketing blargh is going down the hill. 

Nothing is broken about the API. It works just fine. NVidia's hardware and drivers on the other hand are broken almost beyond repair when it comes to concurrent (parallel) async compute. When even the drivers force disables it, for content switching instead, of course you will see degrading performance. Nvidia panicked and launched Pascal instead of Volta like the plan used to state. We know why now. They were caught of guard with these low level API's.

[i_build_nanosuits]

12 minutes ago, Notional said:

I gain 10-15% FPS across the field from min to max fps.

well...why does the whole internet disagree with you this morning?

 

https://www.computerbase.de/2016-09/deus-ex-mankind-divided-dx12-benchmark/2/

 

 

[vitor_cut]

2 hours ago, i_build_nanosuits said:

Yes they did...it's an AMD game.

Also, it's bullshit...Joker tested it with the latest AMD driver and the radeon card lost nearly 10 FPS by going with DX12 and so does the nvidia card, here:

 

 

 

 

Well Pcper and and Guru of 3D notice the AMD performance gains with Crimson 16.9.1 Dx12 driver

 

 

 

 http://www.pcper.com/news/General-Tech/Adam-Jensen-gets-upgrade-and-so-does-your-AMD-Driver-DX12-Deus-Ex

[vitor_cut]

9 minutes ago, i_build_nanosuits said:

well...why does the whole internet disagree with you this morning?

 

https://www.computerbase.de/2016-09/deus-ex-mankind-divided-dx12-benchmark/2/

 

 

And a 1060 to perform similar to a Fury x at 1080p yeah right!!!

 

And the Cpu what a monster!!

[Valentyn]

37 minutes ago, huilun02 said:

Why is there a pattern with AMD affiliated games always run like arse on PC and have piss poor multi GPU scaling?

Even the 670 and 780Ti aged so much better than their AMD equivalents.

AMD has shitty drivers that try to sabotage older cards. Exclusive black box features that shaft performance among other anti competitive practices.

Its no wonder they can't hold a candle to Nvidia's affordable flagship products.

Is..is this for real?

Oh wait, it's about Gameworks
 

[raphidy]

GG my 1070. But I don't see the point of gaming on a broken api implementation. Better stick with dx11.

[Kinda Bottlenecked]

2 hours ago, i_build_nanosuits said:

what are your concern?!

980ti vs Fury X? same deal...both loose around 10 FPS from going to DX12.

 

Why are you bothered with what I'm concerned about?

[3DOSH]

Why are people so butt hurt that Nvidia loses fps with DX12. This isn't news guys, Nvidia tailores the their drivers for each game and since there is no driver out for Nvidia yet performance is gonna tank if the game has implemented any form of async. Recent  example is DOOM where if you check now with the latest drivers from Nvidia,  pascal cards gain substantial fps in Vulcan over previous drives. Not to mention that patch is still in beta and in the patch notes it states its having problems on few cards where performance digress. 

[grayperview]

There seems to be a lot of conflicting reports about this dx12 implementation....for now i chose to go with what Joker has shown in his video....purely because i can see it running with my own eyes and not just some chart on a website, plus i trust Joker and his testing methods.

 

Sure its only with a 1060 and a 480, but i suspect it will scale across the board.

 

I am confident with a bit of tweaking there will be around 10-15% gains on AMD cards and 5-10% on pascal....Maxwell er dunno about that, may aswell just stick with DX11.

[RagnarokDel]

7 hours ago, djdwosk97 said:

In the US it's actually the cheaper of the two (or was as of last week).

That's only true for the 1060 1050TI 3gb which is a weaker card and is more comparable to a 470 is to the 480.

[79wjd]

4 minutes ago, RagnarokDel said:

That's only true for the 1060 3gb which is a weaker card and is more comparable to a 470 is to the 480.

It was also true for the 6/8gb variants. The cheapest (dual fan, non-reference) 1060s were cheaper than the cheapest (dual fan, non-reference) 480s. Since Pascal is mostly limited by voltages more than temps there really isn't any reason to buy a higher tier card (EVGA ACX, MSI 6GT, etc...) are all good.

 

 

I just check now, and they're priced basically the same. 

 

The cheapest 1060s: 

http://pcpartpicker.com/product/HN648d/msi-geforce-gtx-1060-6gb-gaming-x-video-card-geforce-gtx-1060-6gt-oc

http://pcpartpicker.com/product/wz648d/gigabyte-geforce-gtx-1060-6gb-d5-6g-video-card-gv-n1060d5-6gd

http://pcpartpicker.com/product/RJyxFT/pny-geforce-gtx-1060-6gb-video-card-vcggtx10606pb

http://pcpartpicker.com/product/7RKhP6/gigabyte-geforce-gtx-1060-6gb-windforce-oc-6g-video-card-gv-n1060wf2oc-6gd

 

The cheapest 480s: 

http://pcpartpicker.com/product/BZJkcf/powercolor-radeon-rx-480-8gb-red-devil-video-card-axrx-480-8gbd5-3dhoc

http://pcpartpicker.com/product/W34NnQ/gigabyte-radeon-rx-480-8gb-g1-gaming-video-card-gv-rx480g1-gaming-8gd

[zMeul]

source: http://techreport.com/review/30639/examining-early-directx-12-performance-in-deus-ex-mankind-divided/3

 

look at those frame times .. 

AMD Gaming Evolved game, that supposes to paint AMD in a really good way comes and bites them in the arse 

[Notional]

54 minutes ago, zMeul said:

source: http://techreport.com/review/30639/examining-early-directx-12-performance-in-deus-ex-mankind-divided/3

 

look at those frame times .. 

AMD Gaming Evolved game, that supposes to paint AMD in a really good way comes and bites them in the arse 

Just further proves how incompetent Nixxes has been at DX12. They really need to get their shit together. Game is more than half a year delayed, and they didn't even have DX12 at launch. Weeks later and and it's a poor beta. 

At least hawaii/grenada seems to perform good.

[VagabondWraith]

57 minutes ago, zMeul said:

source: http://techreport.com/review/30639/examining-early-directx-12-performance-in-deus-ex-mankind-divided/3

 

look at those frame times .. 

AMD Gaming Evolved game, that supposes to paint AMD in a really good way comes and bites them in the arse 

Ouch. Sometimes FPS numbers are not what they mean.

[Prysin]

23 minutes ago, VagabondWraith said:

Ouch. Sometimes FPS numbers are not what they mean.

what ive been saying for almost a year. But aslong as Nvidia is ahead, nobody cares about the frame-times.

[cesrai]

2 hours ago, Prysin said:

what ive been saying for almost a year. But aslong as Nvidia is ahead, nobody cares about the frame-times.

There is a reason site like PcPer, Euro Gamer  and others post frame times and other measure isn't there ?  I don't think it's fair generalising and saying no one cares about them as long as Nvidia is in the "lead".

[Prysin]

7 hours ago, cesrai said:

There is a reason site like PcPer, Euro Gamer  and others post frame times and other measure isn't there ?  I don't think it's fair generalising and saying no one cares about them as long as Nvidia is in the "lead".

yes the reason is they cannot be bothered. It takes more time to test frame times aswell as FPS. And most reviewers are just too damn fucking lazy to do it. Or too dumb to understand how this kind of testing works and what the results mean.

[Vash_327]

My 390x beats a 980ti- didn't see that coming at all