Don't think it takes a genius to work out SS is gimped in DX11, it barely flipping works, from the comments in the thread it's far from the pinnacle of stability. The first proper DX11 benchmark didn't display any problems that SS seems to display (Heaven 1.0)
Plus, many people seem to completely ignore where AMD has flat out lied in the past about things (The i7 980x Cinebench result which turned out to be a stock 2500K......), so I wouldn't put it past them to put pressure with this SS and a broken DirectX variant (Likewise with Nvidia tbh, an example would be one of the PhysX benchmarks that would gimp itself halfway through for the Hybrid PhysX users)
It is gimped, I posted this in the Star Swarm thread. They have their motion blur on both versions. This however is only intended for Mantle, but I believe they have it in to show how Mantle can perform beyond limitations of DX.
If you had seen any of Oxide's interviews before they released it, you will have heard about how they intend to use extra features such as their multi render motion blur as an addition to Mantle. They brought out a video showing how DX get's completely crippled by this motion blur.
Is it wrong that they put the motion blur on the DirectX version? No not really, seeing as this was released as a tech demo, more than a benchmark tool. It is specifically highlighting a DX weak area and showing how Mantle can power through it. It's not a bad thing, the only bad thing is people see it as a benchmark tool, when it's absolutely not. This is an engine they intend to ship out to other developers and to develop games of their own. They had spent more time optimising in DX than they did in Mantle, they made a point how it took 1 guy, 1 month to incorporate Mantle into the engine.
Them showing these issues where Mantle storms ahead of directx can only be of benefit to everyone, the more aware people become of certain limitations in directx, the more Microsoft will be pressured into creating a more efficient package for all of us to enjoy.
It's blatantly obvious in the cases where Mantle works (which admittedly is practically never for me) that it's more efficient than Directx. Even in it's infancy.
Is it a game changer? What's a game changer in today's market? In a market where parts are being released with negligible improvements on their predecessors. While it's still not amazing and still incredibly limited in scope, it's still a potential game changer. It's giving quite a lot of hardware more significant boosts than some physical hardware upgrades provide, in today's market. So it's hard to deny it's great, even if in my case I get a loss of performance over any gains. It would be incredibly selfish of me to lambast an early Mantle version, based solely because some people have gains and I don't have any.
As for the whole benchmark arguments, until someone comes out with a proper, reliable and consistent benchmark, people will try and compare the two renders in any way they can. Matt has tried to come up with as close to a reliable benchmark scene as possible, which is still not going to cut it for the long run. Until there is software available to get a reliable and consistent result, people will try every part of every map of every game mode of Battlefield 4, to find the strengths and weaknesses of both renders.
I know people are very opinionated, but can we try to not fill the thread with arguments. Nothing wrong with discussing your opinions, but when a discussion goes on for pages and the same things are being said with every post, you know it's dragged on too long and maybe time to back away from the discussion. Nothing is gained by anyone when that happens.
Sorry for the long post, I try my best not to ramble, but more often than not, I fail miserably.
)
