No DX12 Software is Suitable for Benchmarking *spawn*

seahawk said:
There is a big IF in the last sentence. If you want to to move to Vulkan.. which is far from certain or an obvious need.
Click to expand...

Windows 10 + DX12 GPU is at 41% penetration in the steam hardware survey. DX12 GPU but less than Windows 10 is 33%, DX11 and Vista+ at 12%. I was gonna make the point of removing anything but console/DX12/Vulkan from your engine but it seems DX11 is still too big a share of PC gaming to remove it.
 
fuboi said:
Windows 10 + DX12 GPU is at 41% penetration in the steam hardware survey. DX12 GPU but less than Windows 10 is 33%, DX11 and Vista+ at 12%. I was gonna make the point of removing anything but console/DX12/Vulkan from your engine but it seems DX11 is still too big a share of PC gaming to remove it.
Click to expand...
I wouldn't call 12% much of a big share. Likely people on those systems aren't buying big AAA games anyway. Theoretical Vulkan compatibility at 75%+ for Steam is huge.
 
CarstenS said:
The same marketing spin as with most driver announcements.
These are improvements over a quite old driver revision, not 16 miraculous % more perf over the last driver. And much of it is because of regular optimizations for newly launched titles. It's more a track record where they came from over time since mid-2016 when the 368 driver came out.
Click to expand...
Agreed. Per Nvidia's statement the performance difference is between the 1080 release driver 368.81 and the 378.78.
(1) Figure averages the percentage increase of benchmark numbers in the following: GeForce GTX 1080 at 3840x2160 with launch driver 368.81 vs 378.74 on an Intel Core i7 5930K, 16GB DDR4 using Win10 x64. Ashes of the Singularity, Crazy Preset (46.5, 50.9 or 9%), Tom Clancy's The Division 1.6, Max Settings + 1x SMAA Ultra (31.5, 32.7 or 4%) Hitman, High Settings + High SSAO (50.6, 62.1 or 23%), Rise of the Tomb Raider, Very High + 2x SSAA (20.5, 27.2 or 33%), and Gears of War 4, Ultra Preset (41.2, 45.2 or 10%).
Click to expand...
http://nvidianews.nvidia.com/news/nvidia-announces-gameworks-dx12

Granted the some improvements may have been included in versions version 368.81, Anandtech did show some of the improvements since 368.81.
http://www.anandtech.com/show/11180/the-nvidia-geforce-gtx-1080-ti-review/4
 
Here a quick comparaison. ( GTX1080 WQHD ).. But they dont use the 368.81 release driver of 2016 May release. but the last WQHL. ( at contrario of Nvidia.. Nasty HH )
http://www.guru3d.com/news-story/qu...-66-versus-378-78-directx-12-performance.html

It seems they have iron some bugs on Hitman DX12, other games.. no improvement... But well that was somewhat expected looking at the so old driver they was use for the comparaison... waiting for more test from different reviewers ofc.

Now the good news, is this show a good progression on one year ( even if some are surely due to patch on games .)
 
Wasn't all this GPU driver "tweaking" on the part of IHV's supposed to disappear on DX12 and left in the hands of developers? Thought I read that somewhere but could be wrong.
 
pharma said:
Wasn't all this GPU driver "tweaking" on the part of IHV's supposed to disappear on DX12 and left in the hands of developers? Thought I read that somewhere but could be wrong.
Click to expand...

This dont mean drivers couldnt be bugged ... It is the only game who show an increase on performance across the board with different driver but same version of the game. You have another explanation ? ( the only other one is if the new driver dont render something correctly who increase performance ).
 
Last edited:
pharma said:
Wasn't all this GPU driver "tweaking" on the part of IHV's supposed to disappear on DX12 and left in the hands of developers? Thought I read that somewhere but could be wrong.
Click to expand...
You're right for the most part AFAIK, the only thing I can think of for dx12 that can be optimized are PSO (pipeline state objects). (and for some reason I'm thinking split barriers but I think I'm wrong there)
 
PCWorld retested the GTX 1070 and 1080 during our review process, and you could definitely see the boost in DX12 performance. When the GTX 1080 launched in May 2016, it topped out at 63.3 frames per second in Ashes of the Singularity at high settings and 4K resolution in DirectX 12, with DX11 performance hot on its heels just 5fps behind. With the new drivers, Ashes hits 73.7fps in DX12 with the same settings, while the performance gulf between DX11 and DX12 modes widened to 9.5 frames. Giggity.
Click to expand...
http://www.pcworld.com/article/3178...2-performance-with-new-game-ready-driver.html

Guru3D found similar gains ONLY in Hitman though, as did WCCFTech. So did HOCP. Interestingly, Anandtech and PCGamesHardware show the 980Ti closing the gap to the FuryX considerably @4K. HOCP found a rather large 14% increase in fps of Hitman @144p0, Making it the only game to receive such meaningful boost.

1489379901fkk6C8hkMC_8_1.png


 
Infinisearch said:
You're right for the most part AFAIK, the only thing I can think of for dx12 that can be optimized are PSO (pipeline state objects). (and for some reason I'm thinking split barriers but I think I'm wrong there)
Click to expand...
Async scheduling would be an opportunity. DX12 just presents the streams, hardware/driver takes it from there.
 
pharma said:
Wasn't all this GPU driver "tweaking" on the part of IHV's supposed to disappear on DX12 and left in the hands of developers? Thought I read that somewhere but could be wrong.
Click to expand...
DX12 takes things like memory management out of the hands of the driver, but there are often different ways to program the hardware and some things can be tweaked for individual applications.
 
Further confirmation about what I've been saying regarding Crossfire in DX12, particularly in Square Enix games:

 
ToTTenTranz said:
Further confirmation about what I've been saying regarding Crossfire in DX12, particularly in Square Enix games:

Click to expand...


he should take his own advice, please don't post his videos here, bad enough every one that is pro AMD points to his stuff on other forums......
 
Razor1 said:
he should take his own advice, please don't post his videos here, bad enough every one that is pro AMD points to his stuff on other forums......
Click to expand...
That's not very constructive discussion on the video content which is rather compelling. Or are you dismissing the findings completely as rantings of a fanboy and if so, why?
 
He is using a CrossFired RX 480 to beat down a single 1070, and jump to the conclusion that DX12 is faster on AMD in Tomb Raider, even though this is not the case! I think that pretty much sums up the situation here. If your methods are flawed then your conclusions are also flawed.
 
Some differences?:
  • Tomb Raider has NVidia-exclusive graphics options? VXAO?
  • Tessellation, which by default on AMD is capped
If there's a difference in the graphics settings then there might be a difference in the CPU load associated with those settings...
 
So ur saying that benches using Nvidia gpu are not representative of amd cards regardless of gen of performance?

Enviado desde mi HTC One mediante Tapatalk
 
You must log in or register to reply here.

Similar threads

R
id Software releases Open-Source Wolf3D iPhone port
Replies
13
Views
6K
Ailuros
Ailuros
O
Unreal creator Tim Sweeney: "PCs are good for anything, just not games”
3 4 5
Replies
91
Views
18K
Mobius1aic
M
A
Full transcript of John Carmack's QuakeCon 2005 Keynote
3 4 5
Replies
98
Views
34K
nemisis
N
Back
Top