I think you guys need to realize just how long it takes to put together these results. We're not under any kind of restriction from AMD - in fact I'd rather we were, as an NDA would have meant everyone would have gotten additional time to work on results before any articles were published. The lack of NV cards is solely because I (and presumably everyone else) is neck deep in testing, and comparing Mantle to Direct3D on AMD hardware is going to be the higher priority.But judging how reviews are done (NO NV CARD IN SIGHT) - not entirely baseless. What do you think?
Until Mantle is in the majority of games, you cannot switch to a slow CPU and expect good results.Mantle is looking like it's going to give the crown back to the 290X, although for mGPU the 290X never lost the crown. 10-15% is easily enough to put the 780ti in its place.
I wonder what intel is thinking about this. The 4770k/4930K/4960X are no longer going to be great ideas if all you want to do is game. AMD processors are suddenly viable options, assuming Mantle really takes off (and I think it will).
Until Mantle is in the majority of games, you cannot switch to a slow CPU and expect good results.
One game does not give the crown to a particular card, nor should it be the basis of building a PC, unless that is the only game you play.
People are going to want Mantle, and Devs are going to include it because it will help sell games to every person with a GCN GPU.Mantle is looking like it's going to give the crown back to the 290X, although for mGPU the 290X never lost the crown. 10-15% is easily enough to put the 780ti in its place.
I wonder what intel is thinking about this. The 4770k/4930K/4960X are no longer going to be great ideas if all you want to do is game. AMD processors are suddenly viable options, assuming Mantle really takes off (and I think it will).
So did you try Mantle at all? What are your results? Eh?
Avec i5 750 et HD 7870, je passe de 17,5 fps à 36,5 fps
Left is directX, right is Mantle from me in the BF4 test range.
No wonder the game feels way smoother, I wondered if it was a placebo effect.
Nono... thats purely my conjecture.
But judging how reviews are done (NO NV CARD IN SIGHT) - not entirely baseless. What do you think?
http://www.computerbase.de/artikel/grafikkarten/2014/erste-eindruecke-zu-amds-mantle/2/
Try that one, it includes BF4 SP and MP results as well as Star Swarm results. Multiple cards tested from AMD and Nvidia.
Single Player = ~14% increase at 1080p with i7 4770k. 28% increase in MP with the same PC.
They tested in win7, would have been great to see win8.1
So how about incorporating Mantle in EVERY new game?
Or how about building renderers for existing games, especially for so called CPU pigs?
Nice Computerbase.de actually compair the 290x with the 780ti.
And they even use a "common" cpu, like a i7-4770k.
In single player the 290x they have is 1% faster, than the 780ti.
In multi player the 290x is ~20% faster, than the 780ti.
They also keep the 290x, and run the tests at varius CPU speeds (single player) and Mantle is always faster than DirectX for them.
They have (for a i7-4770k):
2.0 ghz 3 cores (no HTT) (mantle) > 4.4 ghz 4 core + HTT (directX)
Mantle makes a huge differnce.
The bad thing about it... out of hundred or so games I own, it works in one single game.
So how about incorporating Mantle in EVERY new game?
Or how about building renderers for existing games, especially for so called CPU pigs?
Q/A wise... AMD needs an army of new software engineers - years ago!
Who ever says that AMD and NV drivers are of similar QA level is out of his mind.
Hire new soft. engineers, Q/A... clean up, build Mantle in more games, maintain, clean up, more Q/A... and AMD has itself a clear winner.
Granted, that assumes that NV/Intel decide to sit aimlessly, and AMD comes up with acceptable cost to benefit ratio.