No DX12 Software is Suitable for Benchmarking *spawn*

[Kyyla]

It does look nice, but not that nice.

 

[Deleted member 13524]

Why are the Deus Ex DX11 results being brought up on a thread about DX12 benchmarks?

 

[trinibwoy]

Deus Ex: Mankind Divided Benchmarked: What Do You Need To Run It?

Ya need a lot to run this game lol, damn. Runs like a dog with three legs

Why do you say that? A GTX 1060 pulls 60fps at 1080p and medium settings. Mainstream card + mainstream resolution + mainstream settings = good performance in my book.

 

[Razor1]

looking at the shadow performance looks like they could do more optimization, 30% drop is way too high for this type of method, I wouldn't accept anything more than 20% drop for that kinda fidelity.

Anycase, for what the game is, I don't see how they hit the graphics cards so hard, you can't play this game in medium settings, with a midrange card from last gen on 1080p, which is the most common resolution, forces people to upgrade, just like Crysis, Far Cry, and other games did.

Not to say the game is bad, looks to be a great game, system requirements seem ridiculous for the visual fidelity it brings to the table.

 

[ninelven]

Why do you say that? A GTX 1060 pulls 60fps at 1080p and medium settings.

If medium were the 3rd tier, I'd probably agree. But it is 4th tier; there is high, very high, and ultra above it. I wouldn't say it runs like a dog, but it is not exactly a model of efficiency either...

 

[Ethatron]

The setting is mislabeled, it's should have been "Expectation". If you put it on medium it runs fine, the higher you put it the worst it gets.

 

[Silent_Buddha]

Is there also a Low, Very Low, and Ultra Low as well? Otherwise the use of Medium seems a bit moronic as it wouldn't be the Medium setting.

Regards,
SB

 

[DavidGraham]

BF1 Beta DX11 / DX12 Test:

DX12 reduces fps on all hardware.
http://gamegpu.com/action-/-fps-/-tps/battlefield-1-open-beta-test-gpu

 

[BRiT]

What new features and functionality does DX12 on BF1 beta bring?

 

[pharma]

Another review with subpar DX12 results.
http://www.pcgameshardware.de/Battl...chnik-Systemanforderungen-Benchmarks-1206197/

 

[Silent_Buddha]

What new features and functionality does DX12 on BF1 beta bring?

Haven't seen anything in English yet saying what, if any differences exist.

However, there's a warning that pops up when you go to enable it that lets users know that the Dx12 path is still under development. So it's quite likely that the state of Dx12 in the BF1 demo is still unoptimized compared to the Dx11 path.

Regards,
SB

 

[ieldra]

Somewhat OT but does anyone know if I can download the "Beyond 3D Suite" benchmark that is frequently used by techreport and PCGH ?

 

[Rodéric]

Somewhat OT but does anyone know if I can download the "Beyond 3D Suite" benchmark that is frequently used by techreport and PCGH ?

I don't think you can, to stay relevant it needs to stay hidden, or something... Or I'm just wrong ;p

 

[Kyyla]

So there is a beta version of Deus ex DX12 render path live now with atrocious results for Nvidia. The devs say it is unfinished. AMD cards get a nice boost though.

http://www.guru3d.com/articles_page..._graphics_performance_benchmark_review,7.html

 

[Deleted member 13524]

@Kyyla it would be nice to post the website where you got the graphics from. They deserve the clicks for all the work they had:
http://www.guru3d.com/articles_page..._graphics_performance_benchmark_review,7.html

1080p results:

So there is a beta version of Deus ex DX12 render path live now with atrocious results for Nvidia.

Atrocious results for nvidia = nvidia cards losing on average 1 (one) FPS beween DX11 and DX12. The 1070 even gains 1 FPS. Regardless DX11 v. DX12 results at 1440p are pretty much the same within margin of error for Pascal, Maxwell and even Kepler cards.

I mean the GTX 980 was just perfect at 35 FPS in DX11, but DX12 completely screws up the whole experience by lowering to a measly 34 FPS.

 

[DavidGraham]

Actually the drop in fps on NV hardware is atrocious indeed, DSO records a 30fps hit in some scenes, guru3d might not catch that because they are using the benchmark, and they are effectively running the game at medium settings (High is two notches down from Ultra).
http://www.dsogaming.com/news/deus-...vailable-performs-poorly-on-nvidias-hardware/

This is from the developer patch notes:

Known DirectX 12 issue:
– There is a known bug that causes some very high end cards to regress relative to DirectX 11. This bug is being addressed by the development team.

http://wccftech.com/deus-ex-mankind-divided-dx12-patch-out-now/#ixzz4JgLWGXrL

 

[Ike Turner]

In game FPS (not built-in benchmark) is actually tanking on GCN GPUs to...so same situation as before launch...
https://www.computerbase.de/2016-09/deus-ex-mankind-divided-dx12-benchmark/2/

GG Nixxes and AMD... /s

 

[Razor1]

yep 2 weeks wasn't going to be enough

 

[pharma]

I could be wrong but think the patch results are displayed on a single Update page:

The DirectX 12 Patch Update

http://www.guru3d.com/articles_page..._graphics_performance_benchmark_review,8.html

 

[DavidGraham]

So same situation as Warhammerl! Built-in benchmark is showing results that are totally different in gameplay. Also Guru3d is testing with High settings, while computerbase is testing with all settings maxed except Contact Shadows. Never the less, computerbase attacks the built in benchmark as being non representative of anything remotely resembling gameplay..

Also fascinating to see FuryX having a hard time differentiating itself from a RX 480 on a Core i3 in DX11. A behavior that all GPUs from both AMD and NV exhibit but in DX12, whereas 1060=1080, FuryX=480. What a joke!