How can nvidia compete against mantle amd cards?

tybert7

2[H]4U
Joined
Aug 23, 2007
Messages
2,763
Unless I missed something, nvidia will have to rely on the slower directx 11/open gl interfaces for games while mantle gpus get to leverage console ports from popular game engines like frostbite.

Even though I am sure all games will still be designed to run on direct x, engines like frostbite 3 are going to be EXTREMELY popular and support a large number of tripple A titles. And that can't be the only large engine that adopts mantle into the builds, as an activision guy tweeted out (even if he does not look forward to the increased workload).

http://www.neogaf.com/forum/showthread.php?t=688033


And which card would enthusiasts prefer to buy? The one that only gets directx 11 optimizations? or the one that can tap into increased performance with the same or vastly cheaper hardware using mantle for popular games?


What about you nvidia guys on these forums, I know plenty of you prefer nvidia for whatever reason, what will YOU be upgrading to for your next cards?
 
What about you nvidia guys on these forums, I know plenty of you prefer nvidia for whatever reason, what will YOU be upgrading to for your next cards?

Me? Probably Maxwell.

I use Linux so this Mantle stuff is kinda irrelevant to me. As is AMD.
 
if mantle takes off it will put AMD in their own space like PhysX does right now for Nvidia. It bypasses OS things so it can include basically everything every feature that DX, OpenGL, OpenCL, audio stacks you name it at a "metal" level, so, if you got real crafty devs say for example DICe or Crytek I think the results will just be jaw dropping, do not pass it off as nothing, AMD will have the support of millions that are able to take advantage of this, and what do you know, Steambox will be getting major Linux support, and this does not have to rely on "windows" to get its advantages so, I think its going to be a major force especially considering they have a billion dollar dev behind it as well as "possibly" MS/Sony at some point once all the grunt work has been proven.

Nvidia shot themselves in foot by not even trying to get in the consoles, that's free dev support for AMD and GCN which is free dev relations for Mantle as well.
 
Nvidia shot themselves in foot by not even trying to get in the consoles, that's free dev support for AMD and GCN which is free dev relations for Mantle as well.

NVIDIA had no chance without an x86 license. None.

And the chances of Mantle coming to Linux anytime soon are close to zero. I don't think EA even knows what Linux is and it's not like AMD has gotten around to making decent OpenGL drivers after all these years. Expecting Mantle too is just a bit much.
 
This thread is obviously intended to be flame bait, but.... Honestly? I've still got a bad taste in my mouth from the last AMD/ATI card I had my hands on. This 'Mantle' shit doesn't change that fact one bit...
 
My god, the doom and gloom. Nv will be fine. Mantle is an advantage for AMD, how much of one I can't say. However I feel pretty confident when I say Mantle is not a "game over man, game over", flag for Nv. In fact, all Mantle really means to me, is that more games will receive AMD specific optimizations than b4. Hurray, better performance for my AMD cards. But, this is something that Nv has had for a while, and AMD is still here. Why does everyone think that this is so huge a deal and Nv is going to close it's doors and suck forever more because of it?
 
Mantle needs to materialize itself as something more than a PowerPoint slide before it has a major effect in anything.
 
Mantle will only benefit from developers using it. If no one uses it, then it's just an API floating out there not supported.

We still do not know yet the potential performance we will get in the real-world, nor do we know how else developers are going to utilize it and to what potential.

AMD is in a good position with it, since they have silicon on the consoles, and PC. The hardware is in place for it, so that's good, now it's up to the devs. I hope it does well for them. I hold judgement on its benefits until I see real-world results.
 
how many games actually use the Frostbite engines?...similar to CryEngine which looks amazing but nobody uses
 
how many games actually use the Frostbite engines?...similar to CryEngine which looks amazing but nobody uses

Games using Frostbit, scroll down

http://en.wikipedia.org/wiki/Frostbite_(game_engine)

Notably, there are a lot more Frostbite 3 games than 2 or 1. So that's positive for the engine.

/edit

There's a lot more games using CryEngine though, and UnrealEngine dominates. If CryEngine and UnrealEngine got on board, it's be epic.

http://en.wikipedia.org/wiki/CryEngine

http://en.wikipedia.org/wiki/List_of_Unreal_Engine_games
 
there is a lot of games now using cryengine, tons that use unreal(Which is generally openGL based unless I am wrong) and quite a few using frostbite.

Anyways I thought this made some sense to those making quips, Mantle will not need to rely on windows and mantle not coming to Linux, it does need to "come to" it is on a core level, so a dev uses it or does not, it does need extras to be used no different then say Cuda, either they use it or do not, they either use DX/OpenGL or do not, it is on a very low level api so devs or even those making the OS can implement it from the base level its not "activated" like a feature would be say like HBAO would be its quite a different ballgame.

Battlefield, Need for speed, the fifa sports if I recall there is quite a few games using frostbite but it is an EA engine I do not believe any games that are not a part of EA use it, Cryengine however is used by many game companies that hold no direct ties to EA.

http://www.cinemablend.com/games/EA...ision-PC-Titles-Support-Mantle-API-59492.html so yeh 2 of the biggest games publishers out there EA and Activision(who have many subsidiaries) will be supporting it, how many games or to what benefit who knows, but the benefit is they can code directly what they want it to be doing going along with the engine so this probably will open many doors for the devs, anyways.

Anyways here is just a few that Mantle can and will be using maybe on launch maybe after launch.

"What's more is that all of EA's Frostbite 3 published titles will make use of the new API's to bring out the best performance from the GPUs, including but not limited to Mirror's Edge 2, Dragon Age: Inquisition, Need for Speed: Rivals, Plants Vs Zombies: Garden Warfare, Mass Effect 4 (that isn't Mass Effect 4) and the upcoming Star Wars: Battlefront."

also as with what I am saying and others not quite understanding this is in black and white

"Combined with the release of the SteamOS, which bypasses Windows to allow developers to also code closer to the metal using Linux, the upcoming generation of gaming is looking like a really great thing for gamers all the way around. In addition to this, Mantle should actually help bolster SteamOS support considering that DirectX isn't supported in Linux but Mantle is. This means that all the performance gain you get out of Mantle will be available in Valve's upcoming OS.

The only thing that remains to be seen is if support for Mantle will last long enough for a strong adoption rate to take place and certify that developers will no longer have to rely on DirectX to get the most out of the graphical designs and features in next-gen games."
 
I do agree with video moderator from [H] until it gets released who really knows what this entails, but I certainly wouldn`t count it out not by a long shot, they obviously have some major names behind it namely Valve, EA/DICE, Activision to name a few and there is a lot of really smart coders for Linux that can and probably will see this be a major force in the works.
 
"Combined with the release of the SteamOS, which bypasses Windows to allow developers to also code closer to the metal using Linux, the upcoming generation of gaming is looking like a really great thing for gamers all the way around. In addition to this, Mantle should actually help bolster SteamOS support considering that DirectX isn't supported in Linux but Mantle is. This means that all the performance gain you get out of Mantle will be available in Valve's upcoming OS.

The only thing that remains to be seen is if support for Mantle will last long enough for a strong adoption rate to take place and certify that developers will no longer have to rely on DirectX to get the most out of the graphical designs and features in next-gen games."

Plus the PS4 runs FreeBSD so it isn't using DirectX anyway. This should turn into a win all around for AMD if work done on the consoles and Mantle translates into better gaming results for Windows and SteamOS users.
 
"Combined with the release of the SteamOS, which bypasses Windows to allow developers to also code closer to the metal using Linux, the upcoming generation of gaming is looking like a really great thing for gamers all the way around. In addition to this, Mantle should actually help bolster SteamOS support considering that DirectX isn't supported in Linux but Mantle is. This means that all the performance gain you get out of Mantle will be available in Valve's upcoming OS.

The only thing that remains to be seen is if support for Mantle will last long enough for a strong adoption rate to take place and certify that developers will no longer have to rely on DirectX to get the most out of the graphical designs and features in next-gen games."

That guy is clueless and you shouldn't be quoting him. There is no evidence at all that Mantle is supported in Linux, nor even a comment that it could even make it over there eventually. I'm not saying that it won't, of course, I'm just saying that there is no information that I know of that says there is any indication at all that Linux support is targeted, let alone in time for a SteamOS release.

And let's not forget that there is a lot of behind-the-scenes politics in this industry. AMD partners with Valve of course, but their relationship with Microsoft takes huge precedence no doubt.
 
Battlefield is epic to FPS as it's the last one to give 100% PC surport with servers after COD forget who made them who they are.

MOHAA was replaced by COD as the FPS WWII era game and after they turned cold shoulder to the PC gamers with Port to Port and no servers, EA gave us Battlefield BC2 and stole all of the PC gamers away from COD and used it to get to where we are day..

So IMO Frostbite engine is alot bigger in gaming surport then most think as gamers that play this type of game stay with it and don't care about other games as much .. MOAHH i played every day for over a year because of clan battles.
 
How are we even having this conversion when there are literally zero benchmarks out for Mantle and won't be until at least November-December? Yeah, I'm definitely interested in seeing how it plays out, but stop all this nonsense please.
 
How are we even having this conversion when there are literally zero benchmarks out for Mantle and won't be until at least November-December? Yeah, I'm definitely interested in seeing how it plays out, but stop all this nonsense please.

You know what happens when you "ass"u"me" stuff :p
 
Mantle has a lot of potential but this thread is kinda ridiculous. Nvidia excels because they make awesome products, and that won't change anytime soon; Mantle is a value add on AMD's part and I applaud them for it - I think it's a fantastic idea. That said, obviously not every game will use Mantle and they need to push it onto other engines such as UE4 as Brent mentioned. That is still very much in the air and if it does happen, it will take time for the fruit of their labors to appear. Mantle won't change the gaming scene overnight, AMD has a lot of work to do still. (and we don't even have benchmarks for Mantle, either..)

NV will still do NV stuff. That is, make kickass products that we all love. But I am happy that AMD is finally (seemingly..) getting their shit together and learned that value added features work - maybe they took a nod from nvidia's TXAA, CUDA and physx. Nvidia sells cards because of value adds, perhaps AMD can make Mantle a good value add like that.
 
This reminds me of the good ol threads a while back..."What will AMD do now that Nvidia has Physx"
 
Nvidia can compete by... creating their own mantle driver, if worst comes to worst why shouldn't they embrace it too since it will be open implementation in theory?
 
Mantle will kill Nvidia in games where its used. Crossfire will scale to 90-100% across 4 gpus at all times as multi-threading will be possible with additional cpu power at disposal 200+ extra fps possible. Mantle will unleash unused untapped cpu power. The more cores you have the merrier. Its going to destroy SLI. Scaling will be deadly. Betting at least 30-80 more fps in most games on a single gpu. Minimum framerates will be doubled meaning higher-lowest fps. Its going to take over once everyone sees the benchmarks. Mantle is the holy grail. Microsoft might send in the dogs some how though if it becomes too successful ;)
 
Last edited:
Nvidia can compete by... creating their own mantle driver, if worst comes to worst why shouldn't they embrace it too since it will be open implementation in theory?

No, Mantle won't work on NV hardware just like physx doesn't work on AMD hardware. AMD would be idiots to give Mantle away, that is their value add that they're pinning their hopes on - they're leveraging it as a value add to create sales for their brand, not nvidia's brand. Besides which, it was confirmed in subsequent meetings that Mantle is AMD only.
 
My guess is some type of open standard comes out if it's even remotely beneficial and that becomes the one that wins.
 
mantle will most likely work on linux at some point

"At some point" is the kicker. For reference, it was 6 months to a year AFTER in-store release before AMD had linux drivers for their 6000-series Radeons. That includes both FOSS and proprietary drivers. Leave it to AMD to sell hardware with zero driver support for a year.
 
No, Mantle won't work on NV hardware just like physx doesn't work on AMD hardware. AMD would be idiots to give Mantle away, that is their value add that they're pinning their hopes on - they're leveraging it as a value add to create sales for their brand, not nvidia's brand. Besides which, it was confirmed in subsequent meetings that Mantle is AMD only.





Mantle is "Open" API meaning even Nvidia can use it. this is where the two differ, AMD ain't so greedy compared to Nvidia LOL


If you watch the video carefully they will say its "OPEN" again means everyone can use it, but of course who else will know how to take advantage of its full potential but AMD coz they created it xD
 
Having left my tri 7970s behind earlier this year for tri sli I'd say all Nvidia needs to do is keep doing what they are doing currently. Lets face it, at the pace AMD moves with software it will probably take a very, very, very long time to develop and make it work properly if at all. When the temptation comes to even consider their new soundcard, I mean R9-290x I just remind myself of the frustration and strong desire to throw my computer through a wall while running my tri 7970 setup. Who cares if Mantle equates to a few more frames, is it worth the headache? I just keep thinking about the look on Dice's face when their Mantle implodes because the R9-290x becomes self aware and starts to cause BSOD and shiz. Of course EA being the parent company will probably sell a Battlepack with a Mantle patch inside to fix the problem for a low price of $59.99.

Does Activision even matter anymore? The last Activision game I played was Keystone Capers.
 
I'm interested to see what mantle brings to the table. Everything right now is just speculation, I guess we'll have to wait until the NDA lifts for some [H]ard numbers.
 
Mantle is "Open" API meaning even Nvidia can use it. this is where the two differ, AMD ain't so greedy compared to Nvidia LOL


If you watch the video carefully they will say its "OPEN" again means everyone can use it, but of course who else will know how to take advantage of its full potential but AMD coz they created it xD

Put simply, you are wrong. Mantle will not work on non-AMD hardware, this was confirmed in closed door meetings after the GPU14 event (this was mentioned by SKYMTL of hardwarecanucks). It is open in that they will make it open source for developers to use and look at for AMD GCN hardware *only* sometime in 2014. But it is not available for other hardware vendors such as nvidia to use. To be blunt, AMD allowing that would be insanely stupid because Mantle is AMD's value add. To let nvidia leverage their technology would basically be giving nvidia free sales that could possibly be their own - a highly stupid business decision. Then again, AMD *is* known for stupid business decisions - yet, they are not quite that incompetent to allow nvidia to use Mantle. That would be outright completely fucking stupid on AMD's part. But the main point here is, Mantle is not available for anyone to use except for AMD GCN - and like I mentioned, this was confirmed after the GPU14 event behind closed doors in meetings per SKYMTL.

Also, AMD still has a lot of work to pitch Mantle to more engine developers, and we don't know the final performance yet - I hope that Mantle is "all that" and a slice of bread, but we just need to see if that is indeed the case; that remains to be seen.
 
Last edited:
imo, this is a big win for AMD. Initially, console games will probably not need a low level API, but eventually most developers will be using it extensively to get better performance out of the consoles.

But say AMD cards get a 20% boost from Mantle - they're still gonna be in the ballpark of NVidia cards. NVidia will just have to get more price competitive. And AMD will likely get lazy and fall behind NVidia in chip development.

So, major win, but certainly not the end of NVidia.
 
My guess is some type of open standard comes out if it's even remotely beneficial and that becomes the one that wins.

Mantle is the low-level API for both PS4 and XBOne. There won't be any other standard, and most developers will be using it, just like they use low-level APIs for the 360 and PS3.

So many or most games developed for the new consoles will have Mantle optimizations on the PC.
 
I'm interested to see what mantle brings to the table. Everything right now is just speculation, I guess we'll have to wait until the NDA lifts for some [H]ard numbers.

We won't have Mantle numbers until December.
 
"At some point" is the kicker. For reference, it was 6 months to a year AFTER in-store release before AMD had linux drivers for their 6000-series Radeons. That includes both FOSS and proprietary drivers. Leave it to AMD to sell hardware with zero driver support for a year.

i actually think it will be sooner rather than later once steamOS comes out, with mantle being open it leaves the door open for valve to get Mantle working on SteamOS. not just AMD.

getting mantle on SrteamOS leaves for the door open for mantle to show up on other linux OS too.
 
"At some point" is the kicker. For reference, it was 6 months to a year AFTER in-store release before AMD had linux drivers for their 6000-series Radeons. That includes both FOSS and proprietary drivers. Leave it to AMD to sell hardware with zero driver support for a year.

Does nvidia have much better linux support? I recall a pretty public rebuke from linus against nvidia for being "one of the WORST" vendors to work with:

http://www.youtube.com/watch?v=IVpOyKCNZYw
 
i actually think it will be sooner rather than later once steamOS comes out, with mantle being open it leaves the door open for valve to get Mantle working on SteamOS. not just AMD.

But you're going to need driver support. Which is like pulling teeth with AMD.

Does nvidia have much better linux support? I recall a pretty public rebuke from linus against nvidia for being "one of the WORST" vendors to work with:

http://www.youtube.com/watch?v=IVpOyKCNZYw

Ohh yes. Much, much better.
 
Nvidia won't be competing in games that support Mantle, that's pretty much a given. The advantages gained from having a low-level API is just too great for graphics intensive games. So the only real question that remains is how many game engines will actually use it. I personally suspect all the big ones will, and depending on how hard it's to work with( not very, I suspect), even some smaller titles might use it. It's really not all that much work to support another API in a game.

That said, I also don't see how Nvidia could *not* be developing their own Mantle-type API right now. They probably got started the very second AMD made the announcement. It'll take some time obviously, but don't be surprised if they announce within 6-12 months or so.
 
Back
Top