[Steam Blog] Total War: Warhammer - pixel-edge MSAA, 64-bit, Compute Shaders and DX12

[Meekers]

I buy almost every Total War game but I wait until a steam sale. The games usually need a few months of patches.

 

[Pottuvoi]

I don't get it why developers don't go as much as possible on the compute shaders side since it seems to be a better approach on doing things and why do gamers find this as a bad thing, more so since using compute you can do complex physics simulations and even having huge numbers and/(or) complex AIs?

Compute shaders aren't always as fast as pixel shaders or may need a drastic rethinking on how you do things.

From their wording I'm quite sure that they remade the simulation part for compute shaders, but rendering is still done trough ROPs and pixel shaders. (I could be wrong though.)

 

[sweetusernames]

Compute shaders aren't always as fast as pixel shaders or may need a drastic rethinking on how you do things.

From their wording I'm quite sure that they remade the simulation part for compute shaders, but rendering is still done trough ROPs and pixel shaders. (I could be wrong though.)

Indeed, it does need a drastic rethinking. In fact, I think Ashes of the Singularity has a completely different rendering pipeline from other games, I'm not sure on the specifics though (http://oxidegames.com/2016/03/19/ob...-film-rendering-2-decades-later-in-real-time/)

Perhaps maybe in the future we can have games that are almost entirely powered by compute shaders.

It's somewhat disappointing imo to see the Fury X's massive 8.6 TFLOPS (which I believe a heavily OCed 980 Ti can reach, not sure, but at that point the 980 Ti is pulling way more power than Fury X) not really translating to much in games.

Hopefully this doesn't leave Nvidia in the dust if there will be game engines that are heavily skewed towards AMDs compute advantage.

 

[Headfoot]

finally more multithreading! Already pre-ordered. Total War is one of my favorite franchises and this one is looking pretty awesome

 

[xthetenth]

Looks great, people playing seem to love it.

And here I am with work that I've got to get done by 9 AM tomorrow already looking like it'll drag on into the night and the best part is the 9 AM is a hard limit because that's when I'm getting vision correction surgery, so let's hear it for my timing.

 

[Mopetar]

It's somewhat disappointing imo to see the Fury X's massive 8.6 TFLOPS (which I believe a heavily OCed 980 Ti can reach, not sure, but at that point the 980 Ti is pulling way more power than Fury X) not really translating to much in games.

It's always been a matter of utilization for AMD. NV cards have fewer SPs, but they tend to be more powerful, or in the case of Pascal capable of operating at vastly higher clock rates.

Fury has more theoretical compete, but it's more difficult to completely saturate all of its SPs. There was previous discussion on AMD implementing technology (based on a patent) to dynamically increase clock rates for a CU when not all of the resources were being used, but no one was sure if this was something that would make its way into Polaris.

DX12 also helps alleviate some of these problems as extra SPs can be used for async compute purposes which should at least keep the most of the SPs fed, but I think we're still over a year off from seeing games that are designed from the ground up for DX12 to take the most advantage of that hardware.

Hopefully this doesn't leave Nvidia in the dust if there will be game engines that are heavily skewed towards AMDs compute advantage.

It won't. By the time DX12 becomes mainstream, Volta will be ready. If they haven't managed to create an architecture to harness DX12 to its full potential by then that's their problem, but they've got too many talented engineers to seriously consider that as likely without some strong evidence. It probably just means that Pascal is not going to age well, but that's a different matter.

Game engines will probably skew AMD anyways though (unless NV pays a company to use Gameworks) as a lot of PC games are console ports and the consoles will all be using some form of GCN-based graphics so AMD gets a free boost from that alone.

 

[airfathaaaaa]

Indeed, it does need a drastic rethinking. In fact, I think Ashes of the Singularity has a completely different rendering pipeline from other games, I'm not sure on the specifics though (http://oxidegames.com/2016/03/19/ob...-film-rendering-2-decades-later-in-real-time/)

Perhaps maybe in the future we can have games that are almost entirely powered by compute shaders.

It's somewhat disappointing imo to see the Fury X's massive 8.6 TFLOPS (which I believe a heavily OCed 980 Ti can reach, not sure, but at that point the 980 Ti is pulling way more power than Fury X) not really translating to much in games.

Hopefully this doesn't leave Nvidia in the dust if there will be game engines that are heavily skewed towards AMDs compute advantage.

star citizen is massively based on compute shaders (as far as we know its the only one so far)

 

[Bacon1]

SemiAccurate got some info - http://semiaccurate.com/2016/05/23/look-total-war-warhammer/

 

[Headfoot]

TW:WH to support Explicit Multiadapter: http://semiaccurate.com/2016/05/23/look-total-war-warhammer/warhammer-amd-multigpu/

 

[antihelten]

star citizen is massively based on compute shaders (as far as we know its the only one so far)

Huh, where are you getting this from? As far as I know Star Citizen makes no more use of compute shaders than CryEngine normally does (which is basically just for lighting to my knowledge, so a fair bit, but hardly massive)

They may have plans to do so in the future, when they port it to DX12, but then again they have a lot of plans.

 

[Pottuvoi]

Indeed, it does need a drastic rethinking. In fact, I think Ashes of the Singularity has a completely different rendering pipeline from other games, I'm not sure on the specifics though (http://oxidegames.com/2016/03/19/ob...-film-rendering-2-decades-later-in-real-time/)

Perhaps maybe in the future we can have games that are almost entirely powered by compute shaders.

Baiscally, they calculate lighting into texture atlas, then render objects to screen using very simple forward rendering pass. (Single texture?)
Not sure if they use pixelshaders to calculate lighting or compute pass.
Has ability to decouple shading from resolution and framerate.

We are seeing first fully compute renderers on console space, so I'm sure that we will see similar on PC as well. (Dreams)

 

[airfathaaaaa]

Huh, where are you getting this from? As far as I know Star Citizen makes no more use of compute shaders than CryEngine normally does (which is basically just for lighting to my knowledge, so a fair bit, but hardly massive)

They may have plans to do so in the future, when they port it to DX12, but then again they have a lot of plans.

well tbh at this stage its not cryengine anymore...

there is a interview with them some 10 months ago that they said they just started to pass whatever they can into compute to have as much "space" on the other pipelines so that they can improve a.i and immersion
we know that so far the ship models and player models are all passing through compute

 

[antihelten]

well tbh at this stage its not cryengine anymore...

there is a interview with them some 10 months ago that they said they just started to pass whatever they can into compute to have as much "space" on the other pipelines so that they can improve a.i and immersion

So as I said they are planning to do it, but aren't currently doing so (any more than Cryengine does as standard anyway)

we know that so far the ship models and player models are all passing through compute

Got a source/quote for that?