Look at the post I was quoting...and then point the finger again.
*looks at post* Nope, still pointing it at you.
So you can present evidence for the "complete and elegant" mantle API?
Good...show me! ()
Tried of empty PR and posts with no data to back up mantle claims...glad to hear you will change that now!
I await your next post filled with evidence and data in excitement!
You also have to recognize that games will be built around the capabilities of non Mantle 1st, and Mantle will be a bonus. Assuming that developers want their games played on the majority of PC gaming machines. Rather than performance improvements, you may see some visual improvements, much like PhysX does (probably different improvements).
I do not think you understood what I meant.It's a one-time low cost to implement Mantle into a game engine.
http://www.youtube.com/watch?feature=player_detailpage&v=QIWyf8Hyjbg#t=1223
He says it's basically 600 lines of code to implement the entirety of Mantle into their engine. It's already done and games coming out of Oxide (and all Frostbite engine games presumably) will just run faster on otherwise much slower AMD hardware - they don't have to develop specifically for it to get some of the benefits.
This will presumably be the same for any other CPU limited game. It's hard to say what the benefit will be for games more dependent on GPU power but there will be benefits simply due to being closer to the metal and simply by removing the CPU limit anyway.
Once Mantle is coded into a game engine that's it, just free performance in every game thereafter with basically no extra effort involved.
The question is, are we really draw call limited? It doesn't appear so..
Most games don't require anywhere near that many draw calls. In fact, the only type of games I can see that could even utilize that many draw calls are the massive strategy games. For everything else, under 10K easily suffices, and since the next gen consoles max out around 20K draw calls, then why should the industry jump through hoops and expand code bases for the small percentage of GCN users to dramatically increase the draw call limit?
BTW, Civilization 5 on the PC can do 15K draw calls a frame due to DirectX 11 driver multithreading. It's clear that the technology has potential, it just needs to be refined.
And since Microsoft has expressed great interest (they've already begun the work according to them) in lowering the overhead of Direct3D on Windows, then we can say with some confidence that future versions of Direct3D will be much lighter than present.
Mantle comes with some serious drawbacks, that's the problem. If Microsoft never had to care about backward compatibility, then how much better would Direct3D be right now?
But backward compatibility is IMPORTANT, and cannot simply be thrown under the bus. Much like cross architectural compatibility is important.
The question is, are we really draw call limited? It doesn't appear so..
Most games don't require anywhere near that many draw calls. In fact, the only type of games I can see that could even utilize that many draw calls are the massive strategy games. For everything else, under 10K easily suffices, and since the next gen consoles max out around 20K draw calls, then why should the industry jump through hoops and expand code bases for the small percentage of GCN users to dramatically increase the draw call limit?
BTW, Civilization 5 on the PC can do 15K draw calls a frame due to DirectX 11 driver multithreading. It's clear that the technology has potential, it just needs to be refined.
And since Microsoft has expressed great interest (they've already begun the work according to them) in lowering the overhead of Direct3D on Windows, then we can say with some confidence that future versions of Direct3D will be much lighter than present.
Mantle comes with some serious drawbacks, that's the problem. If Microsoft never had to care about backward compatibility, then how much better would Direct3D be right now?
But backward compatibility is IMPORTANT, and cannot simply be thrown under the bus. Much like cross architectural compatibility is important.
Mantle comes with some serious drawbacks, that's the problem. If Microsoft never had to care about backward compatibility, then how much better would Direct3D be right now?
But backward compatibility is IMPORTANT, and cannot simply be thrown under the bus. Much like cross architectural compatibility is important.
What some folks forget when talking about those so called "API wars" was that back then, each game was running on its own engine
Today though, most studios are licensing an engine. Just getting Mantle support for the Unreal Engine would guarantee its support for countless of games.
Just getting Mantle into Frostbite 3, and thus all EA's upcoming games, is a great start already.
And if for example Crytek does support Mantle in the Cryengine, they will have a competetive advantage over the other engine makers...Which will pressure the others to support Mantle too
So you can present evidence for the "complete and elegant" mantle API?
Good...show me! ()
Tried of empty PR and posts with no data to back up mantle claims...glad to hear you will change that now!
I await your next post filled with evidence and data in excitement!
We've covered this same thing many times in this thread, every once in awhile, the trolls come back and bring up the pointless examples of past API wars as if its relevant today.
If major game engines incorporate Mantle, its gonna have a good future, period.
Their first attack was "It's only BF4!! LOL.." then "it's only DICE/EA!!"... apparently that didn't work out so well with many major engines now supporting Mantle. Now its API/Microsoft wars and backwards compatibility (for something still young, they worry about THAT many years down the road?!)!
Lonbjerg if you would actually watch that video you'd realize how silly most of your arguments really are. I could understand if 3rd parties were extolling the virtues of Mantle (or AMD), but this stuff is coming straight from the developers, what source is more credible? Answer: none.
We arent draw call limited, yet we saw dice slides for bf3 and bf4 pointing the enormous effort put into fb2 and 3 to achieve a 3-7k draw call game in typical situations, naming all the tricks needed in the game engine to do so.
I personally would love to stop being cpu limited in bf4, considering i can even set - 20 powertune on a single 7950, be throttled to 501mhz and still be cpu limited in 64p all low (cept lighting quality) at full hd
We are talking before NVIDIA's and Intel's DX 11 driver that support command list right?
That your singlethreaded AMD DX 11 driver has issues dosn't prove that we all need Mantle, it just proves AMD dosn't support command list.
Yep, my friends who span i5 ivys and haswells with nv gpus can also get totally cpu bottlenecked in bf4 when you turn settings down at 1080p.
Sorry that i have to burst your bubble, but i actually play the game and my friends too, some of them had to oc their i5s to 4.5 and only alleviate the problem. On my end, i know i wont even have to oc my 8320 once mantle arrives for bf4.
We are talking before NVIDIA's and Intel's DX 11 driver that support command list right?
That your singlethreaded AMD DX 11 driver has issues dosn't prove that we all need Mantle, it just proves AMD dosn't support command list.
Here is OpenGL doing ~100K drawcalls:
http://on-demand.gputechconf.com/gt...32-Advanced-Scenegraph-Rendering-Pipeline.pdf
So why do we need Mantle?
This is also interesting:
Microsoft said:We’re also working with our ISV and IHV partners on future efforts, including bringing the lightweight runtime and tooling capabilities of the Xbox One Direct3D implementation to Windows, and identifying the next generation of advanced 3D graphics technologies.
Lonbjerg if you would actually watch that video you'd realize how silly most of your arguments really are. I could understand if 3rd parties were extolling the virtues of Mantle (or AMD), but this stuff is coming straight from the developers, what source is more credible? Answer: none.