Quantum Break тест GPU (Gamegpu.com)

[PowerK]

I haven't paid attention to Quantum Break. But this game must be really good considering that I see many people complaining about performance of PC version on various forums.
So, putting DX11 vs. DX12 debate aside, what is the technological element(s) taxing hardware in this game ? Saying "poor optimization" on PC version doesn't explain much..

 

[tviceman]

Even if you compare Fury X performance on DX12 it still cannot beat GTX 980 Ti on DX11 in this game.

My gut hunch is that Nvidia can do DX12 just as well as AMD can but it requires more work on the developers end and (probably) some significant help from Nvidia too. We could actually see a reverse trend of the last few years: instead of DX12 games getting faster on AMD over time, patches and driver releases benefit Nvidia more. To me that makes the most sense as console ports are already coded for GCN and DX12 "mirrors" Mantle. Nvidia is going to AoTS improve itself over time in DX12 titles.

 

[IllogicalGlory]

I haven't paid attention to Quantum Break. But this game must be really good considering that I see many people complaining about performance of PC version on various forums.
So, putting DX11 vs. DX12 debate aside, what is the technological element(s) taxing hardware in this game ? Saying "poor optimization" on PC version doesn't explain much..

It is a very nice looking game and the Xbox version had to sacrifice a large amount of IQ to fit all of its effects and high-quality assets into a 30 FPS budget. There are some really nice special effects and it's using every rendering technique in the book.

http://www.eurogamer.net/articles/digitalfoundry-2016-hands-on-with-quantum-break

It's a bit of a mixed bag overall. At a high resolution, the game would look great, but the amount of antialiasing and upscale tricks starts making it look a little bit "soupy" on the Xbone.

 

[Innokentij]

Ohhh nice they fixed the game! Hope they do the same to deus ex and hitman now, what broken games that run very slow and looks worser then crysis 3.

 

[Phynaz]

Not when its obvious that one vendor is losing massive amounts of performance.

Hmmm...I don't recall you stating this when Nvidia was losing performance after game patches. As a matter of fact I believe your statements were pretty close to the opposite of this post.

Should we go back and check?

 

[IllogicalGlory]

The RX 470 is faring pretty well here. 1060 is only 17.8% faster, where it's 27.8% faster on average according to TPU. With the 480 being 20.7% faster than the 470, also based on TPU's graphs, it might be faster than the 1060 in this game.

Partisan stuff aside, performance looks pretty much fine in this version of the game. The hierarchy is about what you'd expect from an older game, with Kepler making its voice heard, 980 Ti beating Fury X by about 14% at 1080p and losing its advantage somewhat at 1440p and 1080p without scaling. Hopefully a DX12 mode is re-implemented, maybe we'd see AMD close that gap a bit. I was honestly surprised that the Windows 10 store version got as much patching as it did, maybe I'll be surprised again. Still, I'm pretty sure no one cares about this game.

And just for fun, wow:

 

[Unreal123]

The RX 470 is faring pretty well here. 1060 is only 17.8% faster, where it's 27.8% faster on average according to TPU. With the 480 being 20.7% faster than the 470, also based on TPU's graphs, it might be faster than the 1060 in this game.

Partisan stuff aside, performance looks pretty much fine in this version of the game. The hierarchy is about what you'd expect from an older game, with Kepler making its voice heard, 980 Ti beating Fury X by about 14% at 1080p and losing its advantage somewhat at 1440p and 1080p without scaling. Hopefully a DX12 mode is re-implemented, maybe we'd see AMD close that gap a bit. I was honestly surprised that the Windows 10 store version got as much patching as it did, maybe I'll be surprised again. Still, I'm pretty sure no one cares about this game.

And just for fun, wow:

Some poster will only care when it is running slower on Nvidia and no matter how much crap the port is. I am 100% sure if GTX 980 Ti was slower then Fury X in this game then some other Nvidia counter parts were slower then this topic would had around 6-9 pages of explanation that how much nvidia has gimped maxwell and kepler.

 

[Bacon1]

Hmmm...I don't recall you stating this when Nvidia was losing performance after game patches. As a matter of fact I believe your statements were pretty close to the opposite of this post.

Should we go back and check?

When I have I ever cheered on gamers losing performance? Please do find any quotes from me where I've done so. I'm all for gamers getting performance increases not decreases. If a game runs better in DX11 than DX12 for someone, then run it in DX11. There is no IQ difference so why would you ever want to run it slower?

 

[Bacon1]

This isn't really an apples to apples comparison like you'd like to portray. AotS is a Radeon optimized title.. The Nitrous engine is specifically optimized for GCN hardware (similar to how UE4 is highly optimized for NVidia) to a much higher degree than other engines. So basically, AMD already gets a leg up over other IHVs courtesy of Oxide's efforts, and this includes DX11.

Why do people keep repeating this? Oxide has said they worked closer with Nvidia than AMD during the development of Nitrous engine and AOTS. Please create a new thread and show any proof of that in it.

 

[Jaydip]

980Ti is one of the best gpus ever released by NV

 

[showb1z]

Why does anyone give a crap about day 1 benchmarks anyway, let alone pre-release. There are always issues to be worked out.
For QB in particular, it took Remedy and Microsoft 2 months to fix up the UWP version to acceptable standards. In the end performance was alright, considering the impressive visuals.

They better hold their promise and update the UWP version though. I really enjoyed this game and want to do a second run with different choices.

 

[Carfax83]

Why do people keep repeating this?

Because it's true, and it's been known since 2013 that the Nitrous Engine would support Mantle. AMD also has this on their website:

TUNED FOR RADEON GRAPHICS

• Ashes of the Singularity™ by Stardock and Oxide Games
• Total War™: WARHAMMER® by Creative Assembly
• Battlezone™ VR by Rebellion
• Deus Ex: Mankind Divided™ by Eidos-Montréal
• Nitrous Engine by Oxide Games

Source

What more proof do you need? Unless you believe that Mantle doesn't equal Radeon optimized? And by the way, I don't think there's anything intrinsically wrong with an engine supporting various low level features for an architecture. NVidia got UE4 after all..

Oxide has said they worked closer with Nvidia than AMD during the development of Nitrous engine and AOTS..

No they didn't. Stardock, the publisher for AotS, was a marketing partner for AMD, and Nitrous Engine had deep Radeon optimizations from the get go courtesy of Mantle.. Oxide collaborated with both AMD and NVidia during development. The main reason for NVidia's direct involvement was because Maxwell didn't support dynamic asynchronous compute, and as a result, asynchronous compute needed to be turned off completely for NVidia otherwise Maxwell would incur a large performance penalty.

That's what was driving NVidia to collaborate with Oxide, because AMD was so dominant when reviewers first started publishing results for the beta.

 

[AtenRa]

Because it's true, and it's been known since 2013 that the Nitrous Engine would support Mantle. AMD also has this on their website:


Source

What more proof do you need? Unless you believe that Mantle doesn't equal Radeon optimized?

FronstBite 3 (BF4) supports Mantle but the game was both optimized for AMD and NVIDIA. Just because it supports Mantle doesnt make the engine AMD optimized only. BF4 is very NVIDIA optimized from the start in DX-11 mode.
Not to mention that Oxide themselves have specifically talked about NVIDIAs optimizations to their engine with the help of NVIDIA sending them code to be used directly in the Engine.

 

[Piroko]

I don't think you've been reading what I've been saying. I never said Remedy couldn't do DX12 at any time. What I've been saying, is that DX12 should only be used by advanced developers that have the skills to implement it properly.

Veto. Programmers need to learn how to write for a new API and it's obvious that we're seeing that learning process currently. We would have never left DX7 behind us with your mentality. Or single threaded game engines.

The huge gains Nvidia is now getting under DX12 in AoTS completely disproves and puts to rest the theory that Pascal can't do DX12. It looks like games built with good all around DX12 support can net similar gains on either AMD or Nvidia.

I agree. Eventually we'll see game engines that simply outscale their DX11 branch in all useful scenarios. And judging by the amount of DX12 ports that have been released already I'd say this will go a lot faster than the DX9 -> DX11 transition.

 

[PhonakV30]

No they didn't. Stardock, the publisher for AotS, was a marketing partner for AMD, and Nitrous Engine had deep Radeon optimizations from the get go courtesy of Mantle.. Oxide collaborated with both AMD and NVidia during development. The main reason for NVidia's direct involvement was because Maxwell didn't support dynamic asynchronous compute, and as a result, asynchronous compute needed to be turned off completely for NVidia otherwise Maxwell would incur a large performance penalty.

That's what was driving NVidia to collaborate with Oxide, because AMD was so dominant when reviewers first started publishing results for the beta.

on other hands, They lie? oh Please Stop! What a hallucination theory!

 

[Carfax83]

FronstBite 3 (BF4) supports Mantle but the game was both optimized for AMD and NVIDIA. Just because it supports Mantle doesnt make the engine AMD optimized only. BF4 is very NVIDIA optimized from the start in DX-11 mode.
Not to mention that Oxide themselves have specifically talked about NVIDIAs optimizations to their engine with the help of NVIDIA sending them code to be used directly in the Engine.

Never claimed that AotS was optimized only for AMD at any time, so I don't even know what you're talking about I said that it was Radeon optimized, which means that it has specific low level optimizations for Radeon GPUs. This comes from the Mantle optimization that Oxide programmed into the engine. And the extent of NVidia specific optimizations was primarily to turn off asynchronous compute..

 

[Carfax83]

Veto. Programmers need to learn how to write for a new API and it's obvious that we're seeing that learning process currently. We would have never left DX7 behind us with your mentality. Or single threaded game engines.

I think we can all agree, that they should learn first, and implement later. This learn as you go "mentality" cost Remedy a lot of credibility with PC gamers due to how badly Quantum Break ran with DX12, as well as sales... I don't think Remedy was ready to utilize DX12 in a shipped title, but Microsoft pushed them into it.

 

[AtenRa]

Never claimed that AotS was optimized only for AMD at any time, so I don't even know what you're talking about I said that it was Radeon optimized, which means that it has specific low level optimizations for Radeon GPUs. This comes from the Mantle optimization that Oxide programmed into the engine. And the extent of NVidia specific optimizations was primarily to turn off asynchronous compute..

As explained, Nitrous Engine is also NVIDIA optimized so i dont know why you only talking about Mantle and Radeon optimizations.

 

[Piroko]

I think we can all agree, that they should learn first, and implement later. This learn as you go "mentality" cost Remedy a lot of credibility with PC gamers due to how badly Quantum Break ran with DX12, as well as sales... I don't think Remedy was ready to utilize DX12 in a shipped title, but Microsoft pushed them into it.

That's a fair argument. I think ROTR and BF4 handled this reasonably well with a proper amount of warnings that this was a test implementation (even though most were ignored by the community...).

 

[Carfax83]

As explained, Nitrous Engine is also NVIDIA optimized so i dont know why you only talking about Mantle and Radeon optimizations.

Right, much like UE4 is AMD optimized . There are degrees of optimization. It's not akin to an on or off switch where an engine is either optimized or not optimized for a particular architecture.. As such, Oxide's Nitrous engine has more GCN optimizations than it does for NVidia hardware. This is a fact, as proven by their Mantle implementation which obviously has GCN specific low level optimization...

So why does NVidia outperform AMD in AotS now, when they were getting their behinds whooped just a few months ago? Well, I think it's mostly because NVidia has made enormous strides in their DX12 drivers over the months, in addition to patches and optimizations from Oxide... As someone that uses NVidia GPUs exclusively and routinely updates to the newest and latest drivers, I can tell you that starting with the 364.51 drivers, there was a big improvement in DX12 titles. And starting with the 372.54 drivers, NVidia made massive improvements in Vulkan performance as well..

So yeah, NVidia was behind the curve when it came to low level API optimization in their drivers, but now they're back in the driver's seat again **pun intended**

 

[AtenRa]

Right, much like UE4 is AMD optimized .

I haven't read anywhere that AMD have sent GCN optimized code to EPIC to include it in UE4 engine. NVIDIA did sent optimized code to be included in the Nitrous engine.

 

[Bryf50]

I don't think you've been reading what I've been saying. I never said Remedy couldn't do DX12 at any time. What I've been saying, is that DX12 should only be used by advanced developers that have the skills to implement it properly. Judging by the benchmarks, it's evident that Remedy are much better with DX11, than they are with DX12..

Remedy has a long history of making high-end graphics engines but they just can't figure out that damned DX12. Seriously, what do you people think DX12 is? I don't know much about graphics apis but I feel like 90% of the people who talk about them know even less.

 

[Spjut]

My gut hunch is that Nvidia can do DX12 just as well as AMD can but it requires more work on the developers end and (probably) some significant help from Nvidia too. We could actually see a reverse trend of the last few years: instead of DX12 games getting faster on AMD over time, patches and driver releases benefit Nvidia more. To me that makes the most sense as console ports are already coded for GCN and DX12 "mirrors" Mantle. Nvidia is going to AoTS improve itself over time in DX12 titles.

AMD also has the advantage that many of the optimizations can be shared between all GCN revisions, whereas Nvidia needs separate optimizations for Kepler to Pascal

 

[moonbogg]

All I know is Forza is a new DX12 game and that stupid thing pretty much uses only a single CPU core. DX12 sucks hard at the moment.

 

[dogen1]

All I know is Forza is a new DX12 game and that stupid thing pretty much uses only a single CPU core. DX12 sucks hard at the moment.

DX12 doesn't mean graphics calls are automatically done in parallel, and it doesn't help making the actual game logic parallelizable.

In the case of forza though, I'd imagine there's something different with the pc version that's causing issues.