What am I supposed to be looking at?
Every card that does 10 and/or 11 does 9.
1/4th of steams user base doesn't have DX11, almost 100% has DX9.
DICE, Crytek, Activision, Square Enix.. and more. I was already optimistic, but now with the Unreal Engine supporting Mantle..
Those are reasons enough.
I believe that Mantle will get significant traction in the market, so don't take this the wrong way; but neither Crytek nor Epic Games have announced any support for Mantle in their engines. A licensee of each engine (CIG and Eidos Montreal) have implemented Mantle themselves, it is not a standard feature in either the CE or the UE yet.
Staggering, indeed. Looks like we might actually see gains well above 20% on average. If so, then I completely understand how Mantle would "ridicule" Titan.
A big reason why DX9 survived for so long is because consoles were built for that kind of hardware. With next-gen consoles coming out, I would be surprised if most games did not start requiring DX11. Remember, PC gamers get console ports, not the other way around.
And you left DX10 out just for the sake of it. It's not because of the market share.
Just as a reminder = DX10 + DX11 = 96.14%
It's because it's way less taxing on hardware and it's been on the market way longer, hence cheaper to develop for, not because you can target 3% more ppl that doesn't have money to upgrade their 7 years old graphics card. You wouldn't even look at them as developer.
Lots of PC games get ported to consoles.
It does not matter what way it goes. Mantle is the api on the consoles also. Just a different name and some minor differences. Nobody says it for marketing reasons. No wonder.
Its the same hardware x86, cgn, dsp, visual connects, probably even same cortex a5 for safety on both consoles.
What amd just made possible is cheap porting and at the same time giving superior performance.
Hey why don't you read what I wrote. I was talking about DX9 and old generation consoles. You love Mantle, we get it, but please actually READ what people are writing.
People continue to confuse draw call overhead with GPU performance.
As long as they get 20% better performance in their games should they care if its from drawcall overhead or some of the 100 other specific gpu related improvements that make their games faster?
Add the stark difference where the problem is in relation to gpu or cpu, is quick becomming an very old way at looking at it
Hey why don't you read what I wrote. I was talking about DX9 and old generation consoles. You love Mantle, we get it, but please actually READ what people are writing.
The real game changer will be when I can run Mantle on my Intel HD xxxx with a GSync monitor giving me fluid render playback.
Point is how are you going to extract 20% better performance out of BF4 if you aren't CPU limited now?
Which gpu specific improvements are those, exactly. You have 100 to pick from, list 20 and their performance benefits.
Unlikely, if Mantle is picked up by all than we'll quickly exceed the capabilities of modern processors, it's human nature.
No it isn't.
DX11 is faster when rendering the same image as DX9, and it has lower draw call overhead as well.
DX10 was a non starter, pointless in our conversation. In this context it should be considered DX9.
What do i care where the performance comes from? As long the performance is there. And they say its there. Its enough for me.
Ask eg the amd engineers, oxide and Johan where the perf comes from. Its a good question.
So, you don't know where, is any, improvements exist and are just picking a number you feel good about then? Gotcha! I can just ignore you from now.
The performance improvements aren't going to come from Mantle itself, except in heavily CPU bound areas, at least, from what I've gathered so far. Where the performance gains will come into play, is where the developers themselves are able to optimize their path. This comes at a large "cost" because developers now have to be better coders. Performance gains are going to be almost entirely up to just how good the developers are at optimizing their code, which they can do already to a point (not as much as Mantle will allow though). So, poorly optimized games aren't going to magically run better and lazy developers won't get some massive gains just "switching" to Mantle.
I also predict a lot influx of memory leaks on rushed games.
One thing developers have control in consoles that they don't have in PCs is memory management - Mantle will allow that.
And, I don't think people understand what a man-month is (or the real impact on software development). If Mantle takes someone 2 man-months (the average amount of work a person can do in 2 months), that increases the development process by 2 months. In situations where you get a 24 month release schedule, giving you maybe 20 months to actually code, an additional 2 months of down time for every employee is huge.
Point is how are you going to extract 20% better performance out of BF4 if you aren't CPU limited now?