Total War: Warhammer DirectX 12 benchmarks

[Racan]

http://gamegpu.com/

 

[Bacon1]

Odd I wonder why MGPU is slower @ 1080, guess its completely CPU bound at that point?

Hmm.. maybe not:

Very odd results. Scaling @ 4k looks semi right, but 1440p / 1080 there is almost no gain

Wonder if they are still working on MGPU support since its DX12 and they have to do it themselves. Maybe Oxide can help them out.

 

[DarkKnightDude]

Kind of funny to see crossfire results in 1080p, they're like less performance than single card lol.

 

[geoxile]

Seeing these results makes you wonder what the GPU marketshare would look like if AMD would optimize their drivers to improve CPU efficiency

 

[Final8ty]

Kind of funny to see crossfire results in 1080p, they're like less performance than single card lol.

That's very common with CPU bottleneck at low resolution.

 

[Silverforce11]

TW: Warhammer isn't that well threaded by design, their AI runs on a separate thread from rendering (which is multi-threaded!), but AFAIK, AI is only single threaded. So in huge battles with a lot of units, the AI thread can (not will, but can) actually be the bottleneck.

Their past games were single threaded, so the AI thread just hammered rendering and so in TW AT & Rome 2, in big battles people's FPS tanked into 20s.

It's a lot better than their other TW titles in optimizations.

 

[Silverforce11]

Tested DX12 patch today myself.

In combat, 20 v 20 stack fighting, FPS held 60, very good. Prior, it was dropping to around 35.

This is at 1080p Medium settings, on a 7950.

On the world map, there's no difference, ~40-50 fps. It seems the DX12 optimizations only affect the real time combat portion of the game engine.

 

[Newbian]

Why is there no current gen cards on it?

Granted the rx 480 just came out and I am not sure when they did the review posted but a lot of sites get them early and could had tested it then and posted it when the nda was lifted much less the 1070 / 1080's.

 

[NTMBK]

Why is there no current gen cards on it?

Granted the rx 480 just came out and I am not sure when they did the review posted but a lot of sites get them early and could had tested it then and posted it when the nda was lifted much less the 1070 / 1080's.

Meh, it's still in beta, I'm sure they'll get around to testing it on a 1080 once they're back in stock.

 

[SPBHM]

no Kepler, GCN 1.0, Polaris and Pascal?

 

[Dygaza]

Guru3d had dx12 bench included in their rx 480 review. It also has 1070 and 1080

http://www.guru3d.com/articles_pages/amd_radeon_r9_rx_480_8gb_review,12.html

 

[Mercennarius]

Why do reviewers still test the 290X instead of the 390X?

 

[cthulhuiscool]

Is there really any visual difference between dx11 and 12? I haven't been able to find a vid that shows any noticeable difference.

 

[Silverforce11]

Is there really any visual difference between dx11 and 12? I haven't been able to find a vid that shows any noticeable difference.

DX12 is not about a visual difference, it's about extra performance.

However, in the longer term, the extra efficiency headroom will allow developers to add more effects.

 

[Zstream]

Is there really any visual difference between dx11 and 12? I haven't been able to find a vid that shows any noticeable difference.

The blood and gore download is cool!

 

[Magee_MC]

Guru3d had dx12 bench included in their rx 480 review. It also has 1070 and 1080

http://www.guru3d.com/articles_pages/amd_radeon_r9_rx_480_8gb_review,12.html

Fury X equal to or above 980Ti, Titan X and 1070 at 1080, 1440 and 4K.

It will be interesting to see if this becomes a trend or if it's an isolated case.

 

[sontin]

Fury X equal to or above 980Ti, Titan X and 1070 at 1080, 1440 and 4K.

It will be interesting to see if this becomes a trend or if it's an isolated case.

Only in the internal benchmark. Ingame Fury X has no chance against a broken DX12 path on Pascal:
https://www.youtube.com/watch?v=uxnFv8cnjZA

 

[Sweepr]

Only in the internal benchmark. Ingame Fury X has no chance against a broken DX12 path on Pascal:
https://www.youtube.com/watch?v=uxnFv8cnjZA

Interesting, so Geforce GTX 1070 does much better than Fury X in-game.

Update 1 for Total War: WARHAMMER will be live on the 30/06 at 2pm BST, and once you’ve downloaded it, our beta version of DX12 will be available to try in-game. Alongside a ton of other optimisation work (see here for details), we’ve been working closely with AMD on implementing DX12, and the results are very promising.

 

[swilli89]

Guess I'm good with my 290X another year

 

[Silverforce11]

Interesting, so Geforce GTX 1070 does much better than Fury X in-game.

You should watch that video again and be-careful how you analyze it. LOL

I just posted my own results and this video proves it.

Campaign world map there's no difference to DX11. Same speed. But in actual battles, the difference is big.

He should have done a side-by-side comparison.

 

[Bacon1]

Only in the internal benchmark. Ingame Fury X has no chance against a broken DX12 path on Pascal:
https://www.youtube.com/watch?v=uxnFv8cnjZA

Wow what is with the stuttering while panning on the 1070, that looks awful. Both Fury X (even while running closer to 34 fps vs 40s) and 1080 are smooth, but the 1070 stutters a lot.

 

[24601]

Interesting, so Geforce GTX 1070 does much better than Fury X in-game.

That's the case in basically every single game released on Windows and when comparing basically every generation of AMD vs Nvidia hardware.

The fact that people still try to dispute the fact is utterly hilarious.

 

[AnandThenMan]

Maybe the 1070/1080 have poor frame times, I wonder what ever happened to this testing.

 

[Silverforce11]

Maybe the 1070/1080 have poor frame times, I wonder what ever happened to this testing.

Devs labeled it a beta DX12 patch. Expect issues.

Like the RotTR beta DX12 patch that so many sites love to use to compare GPUs.

At least in my own gaming, DX12 in TW:W is actually faster than DX11 and that's a win for me.

 

[swilli89]

Maybe the 1070/1080 have poor frame times, I wonder what ever happened to this testing.

I think its kinda pointless to argue anything further, its pretty much over at this point. Have only 200 bucks to spend get a 480, if you have more then buy a 1070 or 1080, there's literally nothing left to discuss on these forums.