I doubt that. They could have just proposed changes to OGL or made extensions if they wanted to.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
Mantle is here
- Thread starter bigbeard86
- Start date
I doubt that. They could have just proposed changes to OGL or made extensions if they wanted to.
Semantics
2[H]4U
- Joined
- May 18, 2010
- Messages
- 2,811
This has very little to do with what directX could do. The extra layer of abstraction used to make directX good is what AMD is skipping over. DirectX is stagnant? Maybe sections of directX like directsound but not D3D... Mantel is hardly offering something new or revolutionary they aren't even offering developers in depth documented SDK for making games which is what DirectX is. So i don't get where you're getting you assumptions from... next thing you're gonna say is AMD had frame pacinging drivers for the betterment of games! not the fact that stuttering esp in crossfire was a common complaint that wasn't found on nvidia's cards. AMD did it for positive press for their APU and they thought they could make a buck off tricking people that they should buy an AMD card because the 5 AAA games in the next 2 years will support mantle or how ever few it's going to be.
whats with the thread crapping?
there are MANY, MANY more than 5 games that have been announced.
The main benefit to mantle is console ports.
Because the language is so close to consoles, it can be ported with a lot less resources than DX
Oxide ported their engine by one dude in his basement in a month.
there are MANY, MANY more than 5 games that have been announced.
The main benefit to mantle is console ports.
Because the language is so close to consoles, it can be ported with a lot less resources than DX
Oxide ported their engine by one dude in his basement in a month.
Tons of extensions have been added to OGL over the years but the end results have been largely unimpressive and it has continued to decline in use for Windows games. It does relatively well in Linux but there is hardly any market for Linux games still so that isn't saying much.
No that too. Developers have been grumbling about it for years but you haven't seen them do much of it in public for political reasons. They need to keep their jobs too you know.
Mantel doesn't need to be revolutionary and I haven't claimed it was so. So please stop whacking on that strawman. It does however offer both nice performance improvements on existing and future hardware vs DX and more flexibility for the developer. SDK aren't required since it uses much of the same tools required for DX and they've already said documentation is coming so what is the problem there? That it isn't available on day 1 of public release or even before it ideally is irritating but not at all a deal breaker.
You sound like NV PR here, I sure hope you're getting paid to spread this FUD.
Semantics
2[H]4U
- Joined
- May 18, 2010
- Messages
- 2,811
...The same dude could port an the engine into directX is a month just as easily creating an engine from scrap is the hard part. A game engine is usually written in something like C. If you code agnostically well enough it's a sorta simple process to just make that work in DirectX or openGL, then you spend time getting to to work specifically on something like ps4 or windows maybe add some specific optimization.
Plus mantel api is nothing like the the api's for consoles, ps4 uses Sony's LibGCM variant and Xbox uses DirectX variant that has a few xbox specific direct calls to parts in the console. It's true developers have used low level access in consoles for years but it doesn't make mantles like consoles enough to mean easier game ports! Porting games would mean exactly the same kind of cost, if not even more because they aren't going to port a game strictly using mantle, but instead mantle + directX.
Also name 5 AAA games announced for release in the next 2 years that will use mantle. AAA games are rated as such due to money put into the project usually just means a major studio.
So your counter argument is, developers have asked for this just I have no proof of this because those developers that are complaining are doing so in secret but i happen to know. Because you know if developers complained about directX their company will fire them because all developers work directly for microsoft and microsoft is a company known for firing people who are critical of microsoft. Then you go on to accuse me of a strawman ending it off calling me a nvidia shill, ie a fallacy in itself ad hominem.
Weird how being a general pessimist and nihilist equates to such things.
Last edited:
There are over a dozen AAA games annouced so far, not 5.
a small sample:
BF4 (obviously)
Star Control (StarSwarm)
Mars (code name, not the final name)
Unannounced stardock game (most likely a SOASE sequel)
Mass Effect
Theif
Star Citizen
Plants Vs Zombies
Dragon Age: Inquisition
Stalker Elite III
10 AAA games right there. There are many more
If you read what i wrote above, i said the "language" is similar to console languages making porting easier. Large portions of code can be directly copied.
It was pointed out by Johan at dice that is was designed intentionally that way.
a small sample:
BF4 (obviously)
Star Control (StarSwarm)
Mars (code name, not the final name)
Unannounced stardock game (most likely a SOASE sequel)
Mass Effect
Theif
Star Citizen
Plants Vs Zombies
Dragon Age: Inquisition
Stalker Elite III
10 AAA games right there. There are many more
If you read what i wrote above, i said the "language" is similar to console languages making porting easier. Large portions of code can be directly copied.
It was pointed out by Johan at dice that is was designed intentionally that way.
Last edited:
Semantics
2[H]4U
- Joined
- May 18, 2010
- Messages
- 2,811
Plants vs Zombies is not one not a 50 dollar purchase, Star Citizen is crowdfounded, your announced title is Unannouced stardock game... Mass Effect? Mass Effect what you mean a teasted game that has no release date or anything except a teaser? Star Control another game that doesn't have anything hell even a teaser. Hell is Star Control your unannounced stardock game?!? Hell if you're counting games that don't even have a title why not the next need for speed or the next mirror edge?
Dragon Age: Inquistion, Check
You oddly forgot Battlefield 4, Check
Theif, Check '
Mars, Check
That's 4 I said 5 for a reason.
Dragon Age: Inquistion, Check
You oddly forgot Battlefield 4, Check
Theif, Check '
Mars, Check
That's 4 I said 5 for a reason.
wonderfield
Supreme [H]ardness
- Joined
- Dec 11, 2011
- Messages
- 7,396
How do you think they should have been managing it? Direct3D was never intended to be a low-level rendering API, and people seem to act as though they've really fallen off the ball for that. For what it is, Microsoft is actually doing a fairly good job with it, and iterating on it at a totally sensible rate.
There are extensions that assist with all of those aspects. Some of them have made it into later core specifications, some of them are well-supported by multiple vendors and some are vendor-specific.
OpenGL is really the sum of its extensions. There's still historical baggage in parts, and it's not specifically aimed to be what Mantle is, but when you're bringing up a 4.x OpenGL renderer, you bring in a good 95% of its functionality through extensions (pulling in the actual functions and using them in place of much of the original deprecated functions). AMD claims, in fact, that they can, and supposedly will, bring most of Mantle's benefits in through GL extensions. Carmack has said similar things along those lines that the limiting factor is simply a matter of how much vendor-specific functionality you're willing to accept in your renderer.
i updated my list
All of those games i listed (10) have been confirmed in develpoment from the publishers themselves.
Crowd funding does not make a game not a AAA.
AAA is production quality. period.
poor wording, I should have said "Unnamed Stardock Game" however, oxide has confirmed its in development. Its most likely going to be a sequel to SIns of a Solar Empire, but thats just my guess,
StarSwarm=Starcontrol, oxide admitted it in their last video.
Plants vs zombies may or may not be a AAA depending on the final game. We shall have to see.
i will drop that one, as its debateable.
I stand firm with 9 on my list, there are more.
If you are one of those "I dont believe anything anyone says unless its in my hands" type of person, then this is a pointless discussion.
All of those games i listed (10) have been confirmed in develpoment from the publishers themselves.
Crowd funding does not make a game not a AAA.
AAA is production quality. period.
poor wording, I should have said "Unnamed Stardock Game" however, oxide has confirmed its in development. Its most likely going to be a sequel to SIns of a Solar Empire, but thats just my guess,
StarSwarm=Starcontrol, oxide admitted it in their last video.
Plants vs zombies may or may not be a AAA depending on the final game. We shall have to see.
i will drop that one, as its debateable.
I stand firm with 9 on my list, there are more.
If you are one of those "I dont believe anything anyone says unless its in my hands" type of person, then this is a pointless discussion.
Last edited:
If anyone is still interested in some benchmarks:
AMD HD 7850 2GB RAM
Core i5-3350 (3.1GHz)
1920x1280 resolution
DX
AVG FPS: 20.96
Average Unit Count: 4065
Maximum Unit Count: 5465
Mantle
Average FPS: 32.55
Average Unit Count: 4137
Maximum Unit Count: 5502
Granted it's a canned demo and all that, but I'm still excited by the results. Now I just need to figure out how to run the demo on my native 1440p resolution. I'll upload those results once I've figured out how to do it.
AMD HD 7850 2GB RAM
Core i5-3350 (3.1GHz)
1920x1280 resolution
DX
AVG FPS: 20.96
Average Unit Count: 4065
Maximum Unit Count: 5465
Mantle
Average FPS: 32.55
Average Unit Count: 4137
Maximum Unit Count: 5502
Granted it's a canned demo and all that, but I'm still excited by the results. Now I just need to figure out how to run the demo on my native 1440p resolution. I'll upload those results once I've figured out how to do it.
wrong thread.
http://hardforum.com/showthread.php?t=1804188&page=2
this just a mantle discussion thread.
But thanks for the benches
nice uplift.
its not as canned as you would think, the AI is dynamic, no 2 runs are the same.
http://hardforum.com/showthread.php?t=1804188&page=2
this just a mantle discussion thread.
But thanks for the benches
nice uplift.
its not as canned as you would think, the AI is dynamic, no 2 runs are the same.
I'm going off of what the DICE dev + others have said on B3D.
If you work for them yes sure they can and will fire you for public comments. If you're a software developer that is affiliated with them then they can certainly make life difficult for you and have done so in the past. They're the Intel of the software world for the PC platform.
Only in your mind.
Flogger23m
Fully [H]
- Joined
- Jun 19, 2009
- Messages
- 16,779
Someone is butthurt. The title of the thread is "Mantle is here". Not "Mantle in Worthless Synthetic Benchmark Made To Make Mantle Look 4 Times More Relevant Than It Actually is ONLY Thread". Get over yourself and stop trolling.
W.T.F....trolling? He was specifically asking about BF4 benchmarks while belittling another benchmark as worthless. There's already a BF4 thread for that but its fine if someone posts them here as well. Its OK if you think that way but some people might be interested so I posted my scores. I wish the mods would clean up the thread crappers.
Last edited:
Flogger23m
Fully [H]
- Joined
- Jun 19, 2009
- Messages
- 16,779
Thanks for proving my point and discrediting your previous post.
This thread is about Mantle. If you can't handle the fact that people are more interested in real world, relevant benchmarks than your scores then post your own thread perhaps.
Thanks for proving my point and discrediting your previous post.
This thread is about Mantle. If you can't handle the fact that people are more interested in real world, relevant benchmarks than your scores then post your own thread perhaps.
LOL...I think people can handle both types of benchmarks. Remember you yourself said this is a general Mantle thread so your being hypocritical. I'm not sure if you're drunk or just looking to argue but you need to calm your tits.
Last edited:
Actually updating it to support newer hardware capabilities and/or removing cruft to improve performance for newer hardware would probably be a good start.
But Mantle isn't all that low level. AMD specifically made it that way so it would benefit future GPU's too without locking it in to a particular GPU architecture. Hell its mostly beneficial for situations where you're CPU limited! MS last update to DX was 11.2 which had minor improvements over DX 11.1 back in what 2012? They don't even have plans to do DX 12 last we heard publicly on the subject.
wonderfield
Supreme [H]ardness
- Joined
- Dec 11, 2011
- Messages
- 7,396
What newer hardware features, specifically?
Semantics
2[H]4U
- Joined
- May 18, 2010
- Messages
- 2,811
Mantle is low level api if it wasn't there would be cpu overhead, and it's locked into a particular gpu arch which is why it only works on a handful of cards. Sure AMD says they are going to expand it and put it on linux etc but what's that time frame? And how reliable is AMD on delivering their history suggests poorly... You also act like Microsoft can unilaterally just make DirectX revisions, they have to work with nvidia, intel and amd when creating DirectX versions as manufactures eventually have to add some hadware support for certain functions. DirectX atm is way more open to others than mantle is, considering mantle is a unilateral attempt to bring back direct access api's to benefit just one set of cards.
FYI, Mantle would have zero developer support if it would stay proprietary in the long run, period.
Dynamic shader allocation....specifically...
which is why it wont work on cards prior to 7xxx
Nvidia also has this capability, which means they could adobt mantle if they wished.
How so? Glide didn't add any CPU overhead vs DX or OGL.
AMD has said it isn't locked into a particular GPU arch. That is why it works fine with GCN 1, 1.1, and 2. It only works with a relative few cards it wasn't practical to try and support older architectures with it.
Probably be a while but this isn't a big deal, Linux is a tiny market vs the Windows market. It makes sense AMD would target the bigger market first.
They own DX, they can do whatever they want with it. They choose to work with the other GPU IHV's but they're not obligated to at all.
DX is no more open than Mantle, they're both proprietary, DX is only widespread because MS also has a OS monopoly. The only open GPU API is OGL.
- Joined
- Feb 23, 2004
- Messages
- 13,149
Mantle has basically extended the life of my processor another 3 years.
Semantics
2[H]4U
- Joined
- May 18, 2010
- Messages
- 2,811
1) glide is a low level api... in fact it's the low level api for graphics cards. Anyone can run glide though a wrapper, guess what the wrapper does, add a cpu overhead. Nvidia could run mantle but they would lose the benefit of a low level api.
2) AMD says it isn't locked, yet it's locked currently to a few cards. Reality doesn't match what they are saying, saying it's supported on GCN is saying it's supported on one base arch, that's locked into one arch.
3)
4)Sure microsoft could do whatever they want but it's not how they historically operated with directX. Companies lobby microsoft, microsoft makes changes as they see fit. Sure it's all decision to microsoft but currently more than mantle is going to offer.
5)No argument there, just saying mantle is far from an open anyone can implement it.
Only if the majority of your games you plan to play for the next 3 years are with mantle support which i find unlikely.
But Mantle isn't being ran through a wrapper, that is why EA/DICE wrote a whole new renderer that uses Mantle. You only use a wrapper to run proprietary API's like Glide on modern DX anyways, they're there for backwards compatibility. Again, Mantle reduces CPU overhead, it doesn't add it!
And those few cards use a different architecture. Just because it doesn't support all or most of their old architectures doesn't mean its locked. CUDA doesn't support all or even most of Nvidia's previous architectures either you know. It'd be silly to expect Nvidia to try and do that anyways.
How they've operated historically changes nothing. Its their API and its proprietary, they do whatever they want with it. Also AMD worked with EA/DICE devs on Mantle so they're clearly open to lobbying too.
Semantics
2[H]4U
- Joined
- May 18, 2010
- Messages
- 2,811
But it would add overhead if nvidia adopts it. The direct calls would have to be ran though something, that something would add the overhead. The point is mantle is only for AMD and mantle with no overhead is only for AMD's GCN cards.
So for nvidia to gain the benefits of a low level api, that would require nvidia releasing essentially glide, hell they own it irrc. So that would mean game developers would have to support 3? 4? api's assuming they want to hit both nvidia's and amd's low level api's plus directX for legacy, and openGL for even more legacy. This seems like a loss for the game industry not a win. DirectX and openGL are great because they add the layer of abstraction allowing them to be hardware agnostic. although opengl is pretty iffy on the hardware agnostic part due to how drivers for opengl shaders are handled which is in part why amd has historically piss poor support of openGL, but i digress.
If Nvidia adopts it they would do so through a driver written in ASM that has been optimized for their video cards. There will be little to no overhead.
Developers and middleware suppliers already typically support many API's. Unreal supports for instance Flash, DX, OGL, OpenAL and I believe HTML. What is more impressive is they do it across many platforms (ie. Android, iOS, Windows). So long as there is a decent working compiler another API isn't a big problem. Edit: you realize all EA/Dice had to change was the renderer for BF4 to make it work with Mantle right? They're not rewriting the whole game, there is no need to!
Last edited:
Everything I've read indicates that Mantle requires a certain hardware feature set not a specific architecture to run, just like every major revision of Directx has required. I think that AMD will have an uphill battle ahead of them to try and get Nvidia to adopt an API created by them but if enough games support it and there is much performance improvement then Nvidia would be foolish to let AMD have it to themselves if they're willing to share(without too many strings attached).
Early results seem to vary from problematic to great, early teething issues are to be expected with a brand new API so hopefully they can fix the issues and developers can get more out of it as they get more familiar with it. It doesn't sound like AMD wants to hoard it so I'm not sure why anyone would be upset about this, the jury is still out IMO but early results are intriguing and I'm interested to see where this goes.
limitedaccess
Supreme [H]ardness
- Joined
- May 10, 2010
- Messages
- 7,592
Unless AMD forfeits all ownership/licensing rights now and in the future for Mantle and makes it completely open (including handing management of it over to an independent or some sort of joint group) there really wouldn't be anything to prevent them from leveraging Mantle for self benefit in the long term.
As such I don't see how it would be anything but a problem in the long term if Nvidia (or Intel) were to help Mantle achieve industry traction.
This is similarly why it wouldn't make sense for AMD to help push adoption for CUDA, PhysX, etc. even if it were offered to them for "free."
Intel and AMD's "relationship" works because each has corresponding leverage(x86 and x86-64). Even then this relationship is hardly what you would call smooth sailing nor was it exactly completely voluntary (dictated by the reality of the situation).
Is there actually any commitment or indication at the moment to make Mantle open and free in the sense that OpenGL is? If not then if you were Nvidia or Intel would you help Mantle achieve larger industry support voluntarily?
As such I don't see how it would be anything but a problem in the long term if Nvidia (or Intel) were to help Mantle achieve industry traction.
This is similarly why it wouldn't make sense for AMD to help push adoption for CUDA, PhysX, etc. even if it were offered to them for "free."
Intel and AMD's "relationship" works because each has corresponding leverage(x86 and x86-64). Even then this relationship is hardly what you would call smooth sailing nor was it exactly completely voluntary (dictated by the reality of the situation).
Is there actually any commitment or indication at the moment to make Mantle open and free in the sense that OpenGL is? If not then if you were Nvidia or Intel would you help Mantle achieve larger industry support voluntarily?
djoye
2[H]4U
- Joined
- Aug 31, 2004
- Messages
- 3,179
The performance improvement looks pretty good for being early drivers but I'm with the others that think that Mantle could go the way of Glide. I give it 5 years. If it's proprietary and it increases development and testing times, which subsequently increases costs, it might not be popular with developers.
Few developers make graphics engines anymore though, the vast majority of games share just a few different ones. That's why I think an API like Mantle, and whatever Nvidia comes up with, is actually viable now, as opposed to in the earlier days when making a simple engine in-house was quite common.
For me personally, what's more interesting about Mantle than these numeric boosts in frames per second is that it gives developers much more CPU power to mess around with in other parts of the game.
For me personally, what's more interesting about Mantle than these numeric boosts in frames per second is that it gives developers much more CPU power to mess around with in other parts of the game.
Last edited:
- Joined
- Jul 29, 2011
- Messages
- 7,460
Is mantle supported with the new Kaveri APU's using the integrated graphics with the current driver? Tried to search this up but literally everything is blocked at work since it has the word "game" in it.
wonderfield
Supreme [H]ardness
- Joined
- Dec 11, 2011
- Messages
- 7,396
Did you mean to say dynamic shader binding? A subset of that functionality has been in place since D3D11.
With the 14.1 beta, yes.
- Joined
- Jul 29, 2011
- Messages
- 7,460
Great, I'm going to buy the 7850k today for my ITX machine. Should work well even if it can't game on maxed settings. I must be getting old because I'm finding I just don't care about the settings so much any more as long as it works and looks somewhat decent.
yes
6xxx series cards do not support that functionality at the hardware level, so they cannot use mantle.
It is the primary hardware requirement for mantle, or so Dice says anyway.
Mantle has just ticked me off big time!!!
I just bought a new motherboard and i5-4760K CPU to use with my new R9 290/(X) cards and a whole new rig, when I could have kept my i7-860 and my ASUS P7P55 WS as my main setup. My fear was the the old system would bottleneck the 290 series.
$100.00 - Case HAF XB EVO
$120.00 - Motherboard AsRock Fatality
$ 220.00 - CPU i5-4760K
$ 80.00 - Ram
That is $520 spent that Mantle COULD have saved me.
When I get a chance I will have to benchmark both machines and see what the real numbers are. So far, from what I have read, seems like Mantle is a winner especially for older CPU's and smoothing out gameplay.
I just bought a new motherboard and i5-4760K CPU to use with my new R9 290/(X) cards and a whole new rig, when I could have kept my i7-860 and my ASUS P7P55 WS as my main setup. My fear was the the old system would bottleneck the 290 series.
$100.00 - Case HAF XB EVO
$120.00 - Motherboard AsRock Fatality
$ 220.00 - CPU i5-4760K
$ 80.00 - Ram
That is $520 spent that Mantle COULD have saved me.
When I get a chance I will have to benchmark both machines and see what the real numbers are. So far, from what I have read, seems like Mantle is a winner especially for older CPU's and smoothing out gameplay.