sonofhendrix
Junior Member
- Dec 8, 2013
- 7
- 0
- 0
Competition is good. NV-haters should remember this. You won't appreciate NV until it's gone from consumer PCs and AMD will have monopoly pricing power on discrete cards.
You guys think Titan Fall could get the mantle treatment? EA is the publisher!
Titanfall uses Valve's Source engine, which as far as I'm aware has no plans to add Mantle.
Even AMD has not hyped Mantle that much.This thread has lost is meaning due to some dreamers.Plz post base on facts.Even Dev has that reality performance improvement can be up to 20% So only discuss on that.Dont post false statement.
DirectX was as fast as Glide, and Glide didnt support 32bit colours.
What we'd see, for examplesay we're rendering shadow maps. There's really not much compute going on. . . . Compute units are basically sitting there being idle. If, at the same time, we are able to do post-processing effectssay maybe even the post-processing from a previous frame, or what we could do in Tomb Raider, [where] we have TressFX hair simulations, which can be quite expensivewe can do that in parallel, in compute, with these other graphics tasks, and effectively, they can become close to zero cost.
If we guessed that maybe only 50% of that compute power was utilized, the theoretical numberand we won't reach that, but in theory, we might be able to get up to 50% better GPU performance from overlapping compute work, if you would be able to find enough compute work to really fill it up.
The 50% figure is a theoretical best-case scenario, but Katsman added, "It seems quite realistic that you would get maybe 20% additional GPU performance out of optimizations like that."
And isn't cross OS or cross platform?
Mantle should really shine the most on their apu chips right? It would bring the lowest frame rate up making it a much smoother experience
nearly 90% mobile discrete market share paired with Haswell ULV ultrabooks.
Yes. The consoles uses 6 relatively weak jaguar cores for bf4 at multiplayer at 60fps. With a Kaveri with 4 far stronger cores the cpu bottleneck is removed from the games even for notebooks. Its a huge change imho. A cheap notebook will play bf4 at eg high at 768 res.
Now still ddr3 is holding the apu back but with ddr4 2015 we enter ps4 gpu capabilities. And on 20nm it will be at probably 35w tdp in a cheap notebook. And with 14nm probably late 2015 everyone and his brother will have access to high end gpu and gaming capabilities in a small portable form factor. Because the notebooks is compatible in speed and api compatible with mantle. Not much room for dgpu then. The dgpu market for notebooks will crash during 2014-2015. Amd will take the small leftover there is from 2016 onwards.
Mantle is in that sense creating an entire new huge market for the games.
Agreed.
From the presentations it's pretty obvious EA sees lots of $$$ in them thar Mantle hills. What EA sees, other developers and publishers will also.
According to Charlie D Nvidia has decided to exit a major market (over the next several years). My guess is it's x86 OEM and consumer graphics business.
i've said this before but it bears repeating.
NV cant really respond until they have a semi decent idea of what kind of numbers mantle will bring to the table.
until there are solid numbers they cant make a move. if mantle is single digits then spending the time and money on a nv version is a waste. if it is in 'ridcule' territory then there is no chance to catch up and they should push for open standards so they can at least utilize the drawcall bypass even if the direct to metal calls wont be available for a year or two.
- 0-8% improvement = do nothing (its within the margins of current delta with amd gpus on twimtbp and other nv favoring games)
- 9-20% = :hmm: try to offload DX overhead onto an on-gpu-die arm core. (techreport podcast was talking about this some weeks ago)
- 21-49% = D: start working on their own api/help ms update DX/gamble on steam OS
- 50+% = :whiste: jump onboard the mantle train and pray for open standard adoption and khronos group control.
the amount of money and time required for a proprietary nv api is going to likely be a significant hit, its not something you can do on a whim or rumors of mantle performance. if you guess wrong the stockholders will be looking for blood.
i've said this before but it bears repeating.
NV cant really respond until they have a semi decent idea of what kind of numbers mantle will bring to the table.
until there are solid numbers they cant make a move. if mantle is single digits then spending the time and money on a nv version is a waste. if it is in 'ridcule' territory then there is no chance to catch up and they should push for open standards so they can at least utilize the drawcall bypass even if the direct to metal calls wont be available for a year or two.
- 0-8% improvement = do nothing (its within the margins of current delta with amd gpus on twimtbp and other nv favoring games)
- 9-20% = :hmm: try to offload DX overhead onto an on-gpu-die arm core. (techreport podcast was talking about this some weeks ago)
- 21-49% = D: start working on their own api/help ms update DX/gamble on steam OS
- 50+% = :whiste: jump onboard the mantle train and pray for open standard adoption and khronos group control.
the amount of money and time required for a proprietary nv api is going to likely be a significant hit, its not something you can do on a whim or rumors of mantle performance. if you guess wrong the stockholders will be looking for blood.
TitanFall is Nvidia game.As i told every one before.Titanfall uses Valve's Source engine, which as far as I'm aware has no plans to add Mantle.
I don't really think that AMD being in next generation consoles provides a benefit to PC game performance. Both of those systems are using their own API which have nothing to do with Mantle or DX for that matter (XB1 is using a highly modified DX which provides direct to hardware access).
Case in point, Assassin's creed IV. This game doesn't have crossfire support. Has faster performance on nvidia cards. Yet, it is on the next generation systems. I just don't see any correlation between PC ports performing better with AMD being in the next gen systems. I also don't see a correlation between next-generation designed games and the PC ports of those games - not unless they're AMD GE or something along those lines. Oh. Another example. NFS: Rivals. This game has terrible graphics on the PC and has a 30 fps framelimit hardcoded. You cannot change this either as the physics engine is completely tied to the framerate. Obviously, NFS: Rivals being on the next generation consoles didn't help it one bit in terms of the PC port.
I hope to be proven wrong, but aside from BF4 - I haven't seen anything interesting at all. Maybe that will change at some point in the future. Now, all that said, I am very interested in seeing how Mantle does. That could be quite a game changer for AMD, but i'm waiting on actual benchmarks before buying into the hype.
AMD will only a bit advantage if it is a port from Xbone or PS4 but it will cost a lot that is why now Developers making game first on PC which much easier to port.
Wow mantle is not even yet and it is not even used yet and how can u tell it is easy when it is costing 10% more to add mantle.No, porting from console to Mantle is much easier than console to DX/OGL. Aaaand it won't cost much either, ~10% of the normal port cost which is buttons.
Well to all you doubters that barley believe an API can improve performance..
I recall many years ago when i owned a 3DFX Voodoo2 12mb card in an AMD K6-2 350mhz machine, and the game was a racing game called Powerslide, it supported Direct X, OpenGL and 3Dfx Glide. I fully recall it having mediocre performance running in both direct x and openGL, but when switching to Glide, it was full on 60fps silky smooth, very noticeable performance gain.