[Digital Foundry] Why DX12 is a gamechanger for PC (and AMD!)

DarkKnightDude

Senior member
Mar 10, 2011
981
44
91
http://www.eurogamer.net/articles/digitalfoundry-2015-why-directx-12-is-a-gamechanger

DX12 is enabled in Windows 10 Preview and its looking promising.

Once we move onto the Mantle and DirectX 12 results, AMD more than redeems itself. There are immense boosts to draw call throughput from start to finish on every processor tested, the largest boost coming from the FX 8350 where the R9 290X receives a frankly monumental boost to performance in the order of 2,000 per cent (!) when single-core DX11 and DX12 scores are compared. Remember, we are only benching one particular element of the rendering process - but regardless, the boost is phenomenal.

The leap in performance applies to both Mantle and DirectX 12, and we happily noted that AMD's DX12 showing actually shows a significant improvement over Mantle (its own API, remember) in every test. Also worthy of comment is that AMD is highly competitive against Nvidia in all areas here - indeed, its lower-end GPUs actually process draw calls faster than their Nvidia equivalents. But the good news is that every piece of hardware we tested sees a boost courtesy of DX12 - we're seeing a far higher utilisation of both CPU and GPU. The figures demonstrate in particular how under-utilised the geometry engines are on our GPUs - what other areas of the graphics hardware are also under-utilised that DX12 could potentially access? The prospects are tantalising.

Of course, just a reminder, the 2000% increase is for a synthetic test. Real world test will be different.


This CPU tidbit is interesting.

But what's curious is that Nvidia's multi-threading support for its DX11 interface actually sees the FX 8350 pump through more draw calls than the i5. Also fascinating is that DX12's multi-thread support sees the vintage 2012 AMD chip trade blows with the i5 - a chip that launched two years later.
 
Last edited:
Feb 19, 2009
10,457
10
76
I like how the author mentioned he noticed the poor AMD DX11 drivers only in known NV games, like CoD & The Crew (both of which I did not play, hence, I thought AMD's drivers are fine heh). Now added to the list, every single GW title since, before they were patched anyway.

But besides this, DX12 is a huge step in the right direction.
 
Aug 11, 2008
10,451
642
126
I like how the author mentioned he noticed the poor AMD DX11 drivers only in known NV games, like CoD & The Crew (both of which I did not play, hence, I thought AMD's drivers are fine heh). Now added to the list, every single GW title since, before they were patched anyway.

But besides this, DX12 is a huge step in the right direction.

Did you not read the article? It gives extensive data on how bad AMD's DX11 driver is in this demo as well.
 

mohit9206

Golden Member
Jul 2, 2013
1,381
511
136
I feel DX12 has been hyped up through the roof. I doubt its going to be that much awesome as much as it has been hyped up to be. It may bring some decent performance improvements but i doubt it's going to be a game changer.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
I feel DX12 has been hyped up through the roof. I doubt its going to be that much awesome as much as it has been hyped up to be. It may bring some decent performance improvements but i doubt it's going to be a game changer.

Pretty much. DX10 had the same mega hype.
 

monstercameron

Diamond Member
Feb 12, 2013
3,818
1
0
I feel DX12 has been hyped up through the roof. I doubt its going to be that much awesome as much as it has been hyped up to be. It may bring some decent performance improvements but i doubt it's going to be a game changer.


Performance is one thing but the introduction of new algorithms and techniques might do wonders for iq.
 

at80eighty

Senior member
Jun 28, 2004
458
3
81
I feel DX12 has been hyped up through the roof. I doubt its going to be that much awesome as much as it has been hyped up to be. It may bring some decent performance improvements but i doubt it's going to be a game changer.

still, i like the idea of overhead being distributed to the IG of a CPU - leave the GPU for the punishing work
boils down to execution though
 
Aug 11, 2008
10,451
642
126
Pretty much. DX10 had the same mega hype.

We can get a pretty good indication from the few Mantle games that came out. I think it will be situation dependent. With certain games and hardware, there could be a big improvement. With other games and hardware, there will be little change.

For instance in DA:I, I saw no improvement at all from mantle, since I have a realtively decent cpu and a now low end HD7770 gpu. If you were cpu limited, the improvement would likely have been more. But yea, just like so many other things in these forums, they hype train is in danger of running out of control.

Maybe at least AMD fans can quit complaining about gameworks crippling their drivers.
 

gamervivek

Senior member
Jan 17, 2011
490
53
91
Amusing, AMD does better with an i5 than an i3(about similarly clocked) and apparently it proves that AMD isn't doing multithreading correctly. :hmm:

As for their benchmarks, there are a few interesting observations:

For both 290X and 970, the multi threaded draw call number becomes about half of the single threaded one, on i3 4130. Same happens with 260X and only 750Ti is unaffected.


On i5 4690, 290X doubles its multi threaded draw call number, while 970 gets halved! While on 8350, there's no change for 290X while 970 doubles its single threaded performance.


Strange, quite strange.
 

Erenhardt

Diamond Member
Dec 1, 2012
3,251
105
101
Back when mantle was introduced, some said devs will not use it because it is too hard to use low level api.
Install base for dx12 is getting quite big, as some old gpus can support it.
The only question is: will devs use it.
 

Glo.

Diamond Member
Apr 25, 2015
5,765
4,670
136
Pretty much. DX10 had the same mega hype.

IF I remember correctly, DX10 was mostly hyped by Microsoft itself, not software engineers all over the world. Which is case of DX12, Vulkan and Mantle itself.
 

futurefields

Diamond Member
Jun 2, 2012
6,471
32
91
Wow, speak of the devil. I was just mentioning yesterday in some threads how AMD's DX11 driver murders CPU's.

"In the graphics card market, AMD is more competitive - but still faces significant challenges from its implacable rival, Nvidia. Thanks to some well-judged price adjustments, the red team has worthy hardware to compete with most of Nvidia's product line (anything up to the £250 GTX 970, basically) but what's become increasingly apparent over the last six months is that AMD's DirectX 11 driver is sub-optimal, particularly relevant for those looking to build a budget PC - an area where AMD offers the best theoretical price/performance level in the market.

We first noticed the issue playing Call of Duty: Advanced Warfare. A Core i3 and i7 run the game in a very similar manner if you have an Nvidia card, but if you're using an AMD GPU, performance collapses whenever the system is drawing a more complex scene. Advanced Warfare isn't a one-off scenario either. Tune your system to favour frame-rate over visual effects and you'll run into a CPU bottleneck on AMD hardware much faster than you will with the Nvidia equivalent. Take a look at this shot of The Crew. The R9 280 is a great piece of hardware and phenomenal value at £130-£140, but pair it with a Core i3 instead of a more capable quad-core processor, and a third of its performance vanishes. Meanwhile, once again, the Nvidia equivalent card holds up effectively."




"In short, Nvidia's driver is processing the same draw calls much more efficiently than its AMD equivalent, maintaining high frame-rates and leaving more CPU resources open to the actual game logic.
 
Last edited:

Flapdrol1337

Golden Member
May 21, 2014
1,677
93
91

futurefields

Diamond Member
Jun 2, 2012
6,471
32
91
I have a better one for you. How about DX11 in Civ BE, Dragon Age Inq, or BF Hardline? It's horrible. Why? Cos there's Mantle for AMD. Devs & AMD optimize for Mantle where its available.

Find me a neutral game where AMD's DX11 driver overhead is the bottleneck, you won't find one, ever.

"The problems with AMD's DirectX 11 driver, amplified here with a draw call specific bench, are put into sharp relief. Firstly, not only is AMD's single-thread performance slower, but the driver is not multi-core aware whatsoever. Not only is Nvidia faster, but it can scale its load over three threads. The Core i5 - the processor with the fastest single-core performance in all of these tests - is the only chip capable of breaking the 1m draw call threshold on AMD hardware, somewhat vindicating our previous contention that enthusiast-level GPUs require an Intel quad-core CPU to get the most out of them. By contrast, Nvidia gets equivalent draw call throughput on every CPU we tested."



Is this Nvidia's fault AMD doesn't have a good driver?
 

JDG1980

Golden Member
Jul 18, 2013
1,663
570
136
In the upcoming years, DX12 should have two effects: better use of multithreading on the CPU side, and less need for game-specific driver optimizations on the GPU side. Both trends should benefit AMD. Of course, this won't save them if they continue to pump out rebranded CPUs and GPUs from 2012. They need to execute in other areas, too.
 

raghu78

Diamond Member
Aug 23, 2012
4,093
1,475
136
Pretty much. DX10 had the same mega hype.

DX12 addresses the huge API overhead and the very high dependence on single thread performance which has been a problem from long before DX10. The first multi core CPUs came in 2005 and DX10 released in late 2006 / early 2007. But even today DX is archaic and is poor in terms of scaleability, API overhead and low level hardware control.

Mantle, Metal, DX12 and Vulkan are all confirming that the direction to make the API as thin (abstraction and overhead) and as efficient and as scaleable and as power efficient as possible is the right one. Remember game engines and games designed in DX12 will perform at a lower overall power as the API is very efficient and much lower CPU power is used which can be saved or used by the GPU portion in SoC and APUs.

http://www.anandtech.com/show/8388/intel-demonstrates-direct3d-12-performance-and-power-improvements

We are also going to see more consistent frametimes as the API bottleneck is alleviated

http://www.anandtech.com/show/8962/the-directx-12-performance-preview-amd-nvidia-star-swarm/8

In a host of ways DX12 is the API which developers have been wanting all along - leaner, faster, more consistent, more efficient (lower hardware power for getting the same performance vs DX11). :thumbsup:
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
IF I remember correctly, DX10 was mostly hyped by Microsoft itself, not software engineers all over the world. Which is case of DX12, Vulkan and Mantle itself.

DX10 was mainly hyped by graphics companies, specially ATI. But quite some from devs as well. MS was the least one to hype it.
 

gamervivek

Senior member
Jan 17, 2011
490
53
91
It seems people are not reading the results which they are getting. Understandable, considering they are not reading it themselves.


A10 7800 doesn't show multithreading increase for any graphics card.

For mutithreading performance, 970 does better than 290X only with 8350(relative to their own single thread performance, 970 does better in single threaded performance of course).

And this is besides the fact that they recommend an i5-2500k for AMD cards instead of the i3-4130 which aren't that far removed in clockspeed. So AMD are actually doing multithreading!
 

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
Back when mantle was introduced, some said devs will not use it because it is too hard to use low level api.
Install base for dx12 is getting quite big, as some old gpus can support it.
The only question is: will devs use it.

I think they will have to since it's the new standard for Windows.
 

NTMBK

Lifer
Nov 14, 2011
10,269
5,134
136
And this is besides the fact that they recommend an i5-2500k for AMD cards instead of the i3-4130 which aren't that far removed in clockspeed. So AMD are actually doing multithreading!

Even if the AMD driver is only single threaded, if it hogs the majority of a CPU core then that doesn't leave an awful lot of resources for the game itself to run on a dual core.
 

gamervivek

Senior member
Jan 17, 2011
490
53
91
If a gpu does better on an i5 than an i3, is its driver single-threaded or multi-threaded? Seriously, this shouldn't be a hard question to answer.

As for their futuremark tests, apparently 290X does more than twice the draw calls on multithreaded test on i5-4690, though it's probable that they might have put in the wrong numbers there.
But look at my first reply here, something funky is going on with i3 with even 970 getting half the draw calls in multithreaded mode.
 
Aug 11, 2008
10,451
642
126
"The problems with AMD's DirectX 11 driver, amplified here with a draw call specific bench, are put into sharp relief. Firstly, not only is AMD's single-thread performance slower, but the driver is not multi-core aware whatsoever. Not only is Nvidia faster, but it can scale its load over three threads. The Core i5 - the processor with the fastest single-core performance in all of these tests - is the only chip capable of breaking the 1m draw call threshold on AMD hardware, somewhat vindicating our previous contention that enthusiast-level GPUs require an Intel quad-core CPU to get the most out of them. By contrast, Nvidia gets equivalent draw call throughput on every CPU we tested."



Is this Nvidia's fault AMD doesn't have a good driver?

Exactly, and still the same old "it is gameworks' fault" excuse is being thrown about. Seems like gameworks has to be brought up in every thread lately, whether it is relevant or not.
 

Hitman928

Diamond Member
Apr 15, 2012
5,622
8,847
136
I honestly don't know what to make of the dx11 results shown, they don't make much sense to me. One thing is clear, thank goodness dx12/Vulcan/Mantle are coming.




compared to



compared to





 
Last edited:
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |