A good conspiracy theory needs a situation where few people would need to know the truth and this fits, but I suspect this doesn't happen. It's much easier to get specific algorithms into a game that are optimized for your hardware than to test levels in a pre-production game, find what works best on your hardware and hope the competitive situation doesn't change before the game ships. Game companies don't like giving a complete game to IHVs early so this testing would need to happen on site at the game developer. There seem to be too many obstacles to this being a practical situation.
No DX12 Software is Suitable for Benchmarking *spawn*
- Thread starter trinibwoy
- Start date
Ike Turner
Veteran
Civilization VI DX12 support press release:
http://www.marketwired.com/press-re...eiers-civilizationr-vi-nasdaq-amd-2142057.htm
Errr since when did Async Compute become a GCN exclusive feature? I guess that because it never really "worked" on Maxwell, AMD's in full PR mode here. But then again NVidia has yet to show it running on Pascal GPU with meaningful results/perf improvements...:-\
No sign of Shader Intrinsic Functions support on any DX12 game yet though... which is disappointing.
http://www.marketwired.com/press-re...eiers-civilizationr-vi-nasdaq-amd-2142057.htm
Errr since when did Async Compute become a GCN exclusive feature? I guess that because it never really "worked" on Maxwell, AMD's in full PR mode here. But then again NVidia has yet to show it running on Pascal GPU with meaningful results/perf improvements...:-\
No sign of Shader Intrinsic Functions support on any DX12 game yet though... which is disappointing.
A good example at the extremes supporting either AMD or Nvidia.
Fallout 4 for Nvidia, Quantum Break for AMD.
It is fair to say they are designed closer to a specific IHV, not saying this is the norm but shows there is influence-collaboration even at the development level that can change priority/focus.
Cheers
As well they should promote the feature. It's not an IHV specific feature but it's definitely a big strength of the GCN architecture that is bringing their performance up to Nvidia and beyond for the comparable perf/$$. Hopefully create some good competition again.
Legal probably would have this interpreted as:
„Asynchronous compute is a DirectX® 12 feature exclusively supported by the Graphics Core Next or Polaris architectures found in many AMD Radeon™ graphics cards.“ -> exclusively as opposed to the earliert VLIW architecture, which, as i have learned here, is also only true, because Cayman is no DX12 card.
renderstate
Newcomer
Fireaxis has always been in very good terms with AMD. Several of their employees are former AMD engineers and researchers. Dan Baker of AotS "fame" is a big AMD supporter (often partecipates to AMD events) and used to work there. This partnership really makes sense given how both companies go way back..
D
Deleted member 13524
Guest
Meaning there's still no hard proof that Pascal has that ability. Or that it gains anything with it.
Ike Turner
Veteran
This is definitely the likely scenario given Nvidia's silence on the matter.
Silent_Buddha
Legend
Using Occam's Razor it's more likely that Quantum Break just defaulted to whatever code was used for the XBO which happens to have AMD hardware. I get the feeling that the decision to make a simultaneous PC release (as opposed to a delayed PC release like Alan Wake) came late in development and the rough nature of the PC port shows that.
And Fallout 4 likely defaulted to whatever path was used for PC development. It's obviously a title that used the PC as the primary development target and platform with Nvidia hardware used during development and then ported to consoles.
IE - unlikely to have been specifically IHV influenced. Just the nature of their development paths.
Regards,
SB
The AO/God Rays/Volumetric lighting/Shadows were all made with Nvidia collaboration going back to earlier generations of Fallout and continuing with Fallout 4.
Quantum Break, it has been out for a long time and no patches to date has been focused on resolving the issues the game has in terms of post processing effects causing heavy performance impact on Nvidia cards, nor for resolving the stability issues.
Notice both are using similar development areas focused for a specific IHV, also there was no attempts for Fallout 3/4 to resolve these causing heavy performance impact on AMD cards.
That is one reason I chose these 2 games; both never patched to work with the other card manufacturer and both causing performance impact with Volumetric lighting-shadows-etc.
But my original point was more to do with how the IHV can influence and possibly change the priority/focus even at the development stage, and these 2 games being the most extreme example of that process (not saying the norm).
Cheers
Last edited:
You can add at least on the Nvidia side, a tons of DX11 games who was run completelly abnormaly on AMD GPU's ( including most of Ubisoft ones ) .. this type of discussion can only turn in round, as this is nothing new, whatever it is DX11 or 12,.. The fact is most games in DX11 on the end ( after the first waves of titles ), have been highly promoted and made in collaboration with Nvidia .. this incude most unreal engine ones, Activsion COD ( that we never seen in any review anyways, even if they sell as crazy ) Assassin Creeds series, Batman series etc ... AMD was the first with DX11 ( Battlefield and some other ), Nvidia have attack really hard then, where a lot of AAA titles was under their wings ( way before gameworks ) . with later release, things seems a bit inverted .. butt just a little bit.
Suddenly, i got the feeling that, some want to say, that every DX11 games where Nvidia was perform att 60 to 120% faster that their counterpart is due to "overhead" of bad driver from AMD .. I dont say that in that was not the case, in some cases ..
So suddenly, AMD is the devil in the low API.
Last edited:
Ike Turner
Veteran
Just to dispel the myth that Remedy didn't bother doing anything to get the game up and running correctly on NVidia hardware..this is from the game's end credit (and no there's no AMD mentions in there just the NVidia engineers)
Two helped with the engine design and 2 others are part of NVidia's developer relation (TWIMTBP). As stated earlier, the game was so far into development as an Xbox One exclusive (the early PC port was decided just a few months before release) that there probably was no way to do better than what we got.
Two helped with the engine design and 2 others are part of NVidia's developer relation (TWIMTBP). As stated earlier, the game was so far into development as an Xbox One exclusive (the early PC port was decided just a few months before release) that there probably was no way to do better than what we got.
Just to clarify my original post and context goes back to this:
It is a good point, and I was showing potentially 2 games that most highlight the influence of the IHV in the game development and its priority/focus.
Cheers
Ah that must explain why the performance is absolutely dire on all Nvidia cards and also the game crashes
See the Quantum Break thread even on this site, or the reviews with decent benchmarking.
So you think AMD would not be involved with it from a console side of things or even PC (Remedy has mentioned AMD being involved along with Nvidia for trying to resolve performance issues)?
And AMD would still had engineers involved with the Fallout games, it does not mean they are getting the engine/post processing effects optimised for their cards though (which they are not in either games case for each IHV).
Cheers
Ike Turner
Veteran
Wut?Ah that must explain why the performance is absolutely dire on all Nvidia cards and also the game crashes
See the Quantum Break thread even on this site, or the reviews with decent benchmarking.
So you think AMD would not be involved with it from a console side of things or even PC?
And AMD would still had engineers involved with the Fallout games, it does not mean they are getting the engine/post processing effects optimised for their cards though (which they are not in either games case for each IHV).
Cheers
So your saying Nvidia had engineers involved but AMD did not.
But somehow on PC the game crashes on all Nvidia cards while also causing dire performance (more so for 970 and 980 and older cards).
This is probably the worst of the recent console ports in terms of general performance for Nvidia cards (970,980 and earlier cards) relative to AMD; Hitman was initially bad but episode 2 did provide performance increases for Nvidia.
Also as I mentioned AMD has been involved as well with Fallout games, but it never meant those post processing effects being changed.
Cheers
Edit:
Here is the thread regarding Quantum Break on PC: https://forum.beyond3d.com/threads/quantum-break-uwp.57771/page-8#post-1916559
Last edited:
It is late and tired here, so not that enthusiastic to find the old historical articles where AMD raised their issues regarding earlier Fallout games, came down to Volumetric lighting/shadows/etc.
However Bethesda do briefly mention Nvidia's further involvement in Fallout 4.
Also shadow detail still hammers AMD cards along with the volumetric lighting, and other integral effects.
And the other side of the coin is Quantum Break with AMD.
Cheers
However Bethesda do briefly mention Nvidia's further involvement in Fallout 4.
https://bethesda.net/?utm_source=Tw...raphics-technology-of-fallout-4/2015/11/04/45
Also shadow detail still hammers AMD cards along with the volumetric lighting, and other integral effects.
And the other side of the coin is Quantum Break with AMD.
Cheers
Not good enough? Why not?
