Hitman DirectX-12 BenchmarksupdateComputerbase

[csbin]

http://www.pcgameshardware.de/Hitman-Spiel-6333/Specials/DirectX-12-Benchmark-Test-1188758/

http://www.computerbase.de/2016-03/...2/2/#diagramm-hitman-mit-directx-12-1920-1080

update：Computerbase

980TI（1380Mhz）vs 390（1010Mhz）


1080P

2560X1440

4K

 

[Good_fella]

Let's see:
AMD have't knew Hitman is coming so no drivers
No Async
Gameworks
Physx
Biased review

Or whatever excuse why AMD is slower in Gaming Evolved/DX12/Async game? :sneaky:

 

[ShintaiDK]

Another DX12 game not showing the AOTS effect, how surprising.

 

[digitaldurandal]

Looks like both cards have okay framerates however for the details I saw on the previous teaser I am not impressed. Basically by the framrates shown in the screenshot it looks like the more you pay the more you get. 980 < Fury < 980ti.

Is this released already?

 

[railven]

I can't keep track which sites are the shill/bias site - is this one of em?

And why no Fury X?

Either way, color me surprised.

EDIT: Wow, the Fury craters at 4K, VRAM issue?

 

[linkgoron]

Let's see:
AMD have't knew Hitman is coming so no drivers
No Async
Gameworks
Physx
Biased review

Or whatever excuse why AMD is slower in Gaming Evolved/DX12/Async game? :sneaky:

I think that the Fury is doing quite well @1440p vs the 980ti.

Either way, I think that we need more scores to actually comment on the benches. This "uploading while benchmarking" thing doesn't have any real value IMO, until the interesting bits have been benched and uploaded.

 

[xthetenth]

It'll be interesting seeing where the 390(X) land. My money's on another "and this is why the AMD fans didn't jump on the Fury in huge numbers" release.

 

[railven]

It'll be interesting seeing where the 390(X) land. My money's on another "and this is why the AMD fans didn't jump on the Fury in huge numbers" release.

It is also another peg in the "console optimizations favor similar GCN versions."



Reading a translation of that article, DX12 seems fundamentally broken and not ready for the lime light. Woof.

 

[Azix]

Best to wait for launch. if dx12 was in the beta sure, but since they are only putting it out with launch I expected a day 1 patch on it.

We give some of us determined DX12-values, however, provided with a large, red flashing question mark: None of these values &#8203;&#8203;is reliable. There is under DX12 other problems. On one of our test systems Hitman crashed with Nvidia GPUs reproducible at Level entry from, on the other hand, Nvidia GPUs ran apparently without problems. However here costs DX12 from WQHD performance - which in turn the Windows Composing Engine could be due to that deprives the game a part of the memory management. With AMD graphics cards to let Hitman start under DirectX 12 only in windowed mode. The construction status of Hitman under Direct X 12 owed we can currently work only unfortunately regular DX11 benchmarks with useful results rather than the certainly highly anticipated DirectX 12 benchmarks. This we will supplement as soon as possible, in addition, we will continue to deal with the DirectX 12 problems. We ask for some patience.

guys should also keep in mind, the fury is still destroying the 980, its competitor. A fury non-x being close to a 980ti is good for any fury buyer. Nano would be similar

 

[Bryf50]

Ah start us off with a nice helping of shit posts. Anyway this is the $500 Fury not the Fury X. Look at the 1440p benchmarks, the 390x is wrecking the 980.

 

[csbin]

update&#65306;

 

[PhonakV30]

Let's see:
AMD have't knew Hitman is coming so no drivers
No Async
Gameworks
Physx
Biased review

Or whatever excuse why AMD is slower in Gaming Evolved/DX12/Async game? :sneaky:

Slower? 390 is faster than 980 in both DX12 / DX11.Biase review?

 

[crisium]

Ah start us off with a nice helping of shit posts. Anyway this is the $500 Fury not the Fury X. Look at the 1440p benchmarks, the 390x is wrecking the 980.

Yup, we have a game where the $330 390 is beating a $500 980 and Shintai and Good_fella are shitposting on AMD. Are the standards of Nvidia fanboys that low now? Do we need the 390 to beat the 980 Ti?

 

[SlowSpyder]

I bought my R9 390 before the promotion by AMD. Contacted Newegg, they gave me a game code. (thanks AMD, Newegg!) Looking forward to playing this one on DX12, game looks great and should fly on my system.

 

[Snafuh]

PCGH.de is a legit website but their I don't like their recent live benchmarking articles. Uploading something just to be the first is not the way to go.

UPDATE: Performance-Probleme mit AMD, Abstürze bei Nvidia gelöst, neue Benchmarks!

They solves the performance and framelock issues on AMD and the crahses on Nvidia. The first benchmarks are not valid anymore.
They will rebench the game and update live. At the moment the chart features the updates R9 390 numbers and the old numbers for other cards. Really nice for anyone who reads the article now...

 

[railven]

PCGH.de is a legit website but their I don't like their recent live benchmarking articles. Uploading something just to be the first is not the way to go.



They solves the performance and framelock issues on AMD and the crahses on Nvidia. The first benchmarks are not valid anymore.

Man are we not Live Tweeting benchmarks reviews? How freaking instant-gratification/click-baiting has our society become?

Are they streaming this, too?

 

[monstercameron]

I bought my R9 390 before the promotion by AMD. Contacted Newegg, they gave me a game code. (thanks AMD, Newegg!) Looking forward to playing this one on DX12, game looks great and should fly on my system.

Didn't you have a 780ti?

 

[.vodka]

Yup, we have a game where the $330 390 is beating a $500 980 and Shintai and Good_fella are shitposting on AMD. Are the standards of Nvidia fanboys that low now? Do we need the 390 to beat the 980 Ti?

This. A conservatively clocked 390 (1010MHz, Hawaii is good for 1100-1150MHz) is wrecking the 980 (mind you, NOT the full Hawaii in the 390x, cut down Hawaii!) and getting dangerously close to a 1380MHz 980Ti. Not a reference 980Ti, an OC'd one, and by a good amount.

Do we need to remind some people that the 390 is the 970's competition, the 390x is the 980's and the Fury/X are the 980Ti's? I seriously can't understand the logic behind these two posts. AMD is destroying nV's offerings in price/performance in almost all these new games, the 980Ti is nV's only relevant card in production. Between AMD's driver work in DX11 games and these new games, their perf/w is jumping by leaps and bounds. GCN just got even more competitive!

I wonder how Fiji would do in this game and the others considering its weird bottlenecks.



Hawaii is seriously the next Tahiti, G80 and R300/50/60. It's insane. Pascal had better get things straight for nV as far as longevity is concerned, Kepler cracked after time, Maxwell is starting to crack, both under the same competition. It's quite mind boggling.

 

[crisium]

Man are we not Live Tweeting benchmarks reviews? How freaking instant-gratification/click-baiting has our society become?

Are they streaming this, too?

It is sad.

Anyway, this partly explains the barking from earlier as the results have changed. Although I am certain even the original results showed a Fury (non X) beating a 980 which at 1080p cannot possibly be twisted to an AMD failure here, but two users interpreted this as such anyway.

So maybe now the 970 will beat the 390 after results are changed. Who knows. The editors need someone to slap their hand away from posting before they are COMPLETE.

 

[SlowSpyder]

Didn't you have a 780ti?

No, bought a reference 7970 at launch, have been using that until a few weeks ago. Have a Sapphire R9 390 now.

 

[ShintaiDK]

PCGH.de is a legit website but their I don't like their recent live benchmarking articles. Uploading something just to be the first is not the way to go.

They solves the performance and framelock issues on AMD and the crahses on Nvidia. The first benchmarks are not valid anymore.
They will rebench the game and update live. At the moment the chart features the updates R9 390 numbers and the old numbers for other cards. Really nice for anyone who reads the article now...

Seems to be odd workarounds for both as well with Framelock and Render Target Reuse . Not exactly giving confidence in DX12 outside consoles.

 

[Headfoot]

Let's see:
AMD have't knew Hitman is coming so no drivers
No Async
Gameworks
Physx
Biased review

Or whatever excuse why AMD is slower in Gaming Evolved/DX12/Async game? :sneaky:

What are you talking about?

The screenshots show the 390 catching up to a 980 Ti. Not even a 390x. It's absolutely destroying the 980. I have no idea where you are getting "AMD is slower." This probably the single best showing for an AMD card relative the competition, ever. At least since the unified shader.

 

[SPBHM]

Seems to be odd workarounds for both as well with Framelock and Render Target Reuse . Not exactly giving confidence in DX12 outside consoles.

early days, wait another year.

 

[ShintaiDK]

What are you talking about?

The screenshots show the 390 catching up to a 980 Ti. Not even a 390x. It's absolutely destroying the 980. I have no idea where you are getting "AMD is slower."

The article is updated with entirely new benches. Framelock disabled on AMD; Render Target Reuse disabled on NVidia.

 

[Snafuh]

Seems to be odd workarounds for both as well with Framelock and Render Target Reuse . Not exactly giving confidence in DX12 outside consoles.

Devs will get used to it and AMD/Nvidia will improve the drivers. At the moment DX12 seems like a checkbox feature leading to more problems than improvements. In the long term DX12 will definitely be an improvement over 11. Sometimes I look at DX code and wonder who came up with this.
But I don't know what Microsoft has in mind for devs who don't want to deal with lower level access. Khronos wants to continue supporting OpenGL but I guess DX devs are stuck with 11.