Quick test: DirectX 12 API Overhead Benchmark results

#5038589

PrMinisterGR

2015-03-26 21:32

Well well... AMD's DX11 CPU efficiency is abysmal. Who would have guessed... 😀

#5038593

Hilbert Hagedoorn

Administrator

2015-03-26 21:40

This thread is about the massive potential DX12 offers, so let's not make this yet another boring and irritating AMD versus Nvidia bashing shall we ?

#5038607

Wagnard

2015-03-26 22:06

Here is another run on the 970 at 1920x1080 this time. (by default its 720p) [SPOILER]https://s3.amazonaws.com/uploads.hipchat.com/297236/1851429/7cKaZVuZoukoIJz/win%2010%201080p%20970.PNG[/SPOILER]

#5038619

Extraordinary

2015-03-26 22:21

^^ You've got one lol

#5038627

Lane

2015-03-26 22:38

Via PCworld, some fun with TitanX : http://www.pcworld.com/article/2900814/tested-directx-12s-potential-performance-leap-is-insane.html http://i503.photobucket.com/albums/e435/lanek66/dx12_api_perf_4770k_titanx-100575302-orig_zpsxfdqg5th.png http://i503.photobucket.com/albums/e435/lanek66/dx12_api_perf_mantle-100575202-orig_zpsezghtsfd.png Something look strange on MT ( whatever it is Nvidia or AMD), depending the driver it seems. anyway this bench is not for compare gpu's, nor driver..but API on a certain similar test.

#5038632

cyclone3d

2015-03-26 22:48

I got ~ 1million draw calls boost by dx11 single threaded when going from 1333 9-9-9-24-2t (20gb/s, 65ns) to 2400 10-12-12-31-1t(36gb/s, 41ns) By multi it was less, ~ 100k calls difference.. Still something 🤓 I see dx12 makes a nice boost

Oooh.. I am gonna have to test this on my x79 system with different RAM speeds. Might as well set up a disk with Win10 as well to be able to do the DX12 tests. And this is exactly why current and even older games seem smoother even if the framerate doesn't go up all that much when using really fast RAM.

#5038641

orky87

2015-03-26 22:58

^^ You've got one lol

You've got a better one which will make it more interesting.

#5038642

lmimmfn

2015-03-26 23:00

This thread is about the massive potential DX12 offers, so let's not make this yet another boring and irritating AMD versus Nvidia bashing shall we ?

Hilbert, any chance you could test with high core count cpu with not great per core performance vs lower core count with great per core performance and with/without HT? e.g i7 4770K vs ancient Xeon 6 core 5660( just an example latest 8 core AMD is a cool alternative ) Point is, it would be good to see if DX12 removes the cpu race for the latest and greatest delivering higher per core performance vs older more slower core performance but higher core count.

#5038643

Extraordinary

2015-03-26 23:01

You've got a better one which will make it more interesting.

I don't have 10 installed on anywhere other than VMware, and will not install it on real hardware while it's full of keyloggers and data collecting stuff, and in a very unfinished state The guy wants results from an FX CPU and has an FX CPU

#5038645

BetA

2015-03-26 23:02

here are some computerbase tests...to roundup the picture.. http://s23.postimg.org/bgbwkrdd7/dhdfghdfg.jpg http://www.computerbase.de/2015-03/directx-12-im-test-vor-mantle-und-amd-vor-nvidia/#diagramm-3dmark-intel-core-i7-4770k

#5038651

harkinsteven

2015-03-26 23:06

Does this need Win10/dx12 drivers to run?

#5038654

orky87

2015-03-26 23:11

I don't have 10 installed on anywhere other than VMware, and will not install it on real hardware while it's full of keyloggers and data collecting stuff, and in a very unfinished state The guy wants results from an FX CPU and has an FX CPU

Yeah, I know, but honestly I'm just as curious to see how much dx12 is lagging behind Mantle with FX cpu. I suppose it'll be similar to FX/Mantle vs Intel/Mantle score in gaming benches if the ratio can be scaled.

#5038662

Rich_Guy

2015-03-26 23:23

Results from both PCWorld and PCPer are in. PCWorld’s results show the difference in API draw call handling between both the Nvidia GeForce GTX Titan X and the AMD Radeon R9 290X. With DX11 the Titan X manages a maximum of 740 thousand drawcalls per second and a whopping 13 million, 419 thousand calls with DX12. In comparison the 290X manages a maximum of 935 thousand drawcalls with DX11, so more than the Titan X to begin with and again an unbelievable 13 million, 474 thousand calls with DX12. So the 290X slightly edges out the Titan X in DX12 drawcall handling.Interestingly enough AMD’s latest drivers have improved DX12 performance to the point where it’s actually a head of Mantle by about 8%. PCPer’s testing is more traditional, showing the frames per second instead of the number of drawcalls per second. The GTX 980 delivers 15.67 FPS in DX12, a massive improvement compared to the 2.75 FPS it manages in DX11. On the other hand AMD’s R9 290X delivers 19.12 frames per second in DX12 a significant jump over the GTX 980. And perhaps what’s more interesting is that the next iteration of 3DMark which will debut with Windows 10 also supports Mantle. And the 290X delivers 20.88 FPS in the same benchmark running Mantle. A 33% performance lead over the GTX 980 running DX11 and a 9% lead over the same 290X running DX11.

290X edging out the Titan X, LOL.

#5038665

orky87

2015-03-26 23:29

290X edging out the Titan X, LOL.

How about original Titan, wonder if it's compute power will have an advantage over Titan X.

#5038670

Redemption80

2015-03-26 23:44

It does look like AMD cards gain alot more than Nvidia, and not just because how poor AMD DX11 performance is, a 290x is around 25% higher in DX12 than a 980. Though, reading the last few pages again it seems the results vary so much you can't actually accurately compare them, not even to identical hardware.

#5038681