Mantle Thoughts

Communities Gaming
1

Hello,

I recently was saw an article about AMD's Mantle API. It got me thinking about its impact on the industry. I think that this will have a negative impact on the industry. This is coming from a person that has two AMD GPUs running in his primary rig.

I believe that Mantle is hurting PC gaming. Mantle just adds another API into the mix. It competes in the same arena that both DirectX and OpenGL compete. The thing about DirectX and OpenGl is that they work on most hardware. NVIDIA, AMD, and Intel. This is good, because it allows developers to not have to use multiple API's to accomplish the same task,

Mantle makes this a problem. Mantle works on Linux and Windows. OpenGL does this too; however, NVIDIA is not going to support Mantle. This is a problem for developers. It is painstaking to optimize video games under any graphics API. To write a game in two APIs just makes the problem worse. I predict that Indie studios will not take advantage of Mantle because it will severely increase development time and cut the possible audience.

What do you guys think?

2

One year ago, there were 0 meaningful low level graphics APIs outside the console space. Now, there are 4 either out there today or in development; Mantle itself, Metal (for iOS), OpenGL Next, and DX12. Inside 3 years, every major PC graphics API will be in the style of Mantle, rather than the style of old OpenGL/D3D.

Mantle has already changed the entire API landscape for the better, even if Mantle itself is no longer of interest in another year or so.

The actual issue you're talking about is fragmentation. That is not really an issue with Mantle, because there is no reason to use Mantle at the exclusion of anything else. It is there if devs want it (and they do seem to want it), anybody else is free to ignore it. The only devs who even have to know about it, really, are the engine programmers, who are definitely willing and able to make Mantle work to its fullest.

3

Mantle is good for all gaming. While mantle will likely die it did something amazing. Made direct x and ogl makers get off their asses and make them low level api. DX12 and OGL next will be low level like mantle. We need those, mantle may not make it in the long run but it was an important thing.

4

nobody is forcing developers to use mantle, but its a huge improvement from DX and OGL because those have been used for years and they need a lot of changes in code to be updated to the point where they will perform like mantle

5

+1

6

The focus on performance is why many non-programmers feel that Mantle is not successful. It's kind of AMD's fault for selling it originally as a performance improvement, because that's the only factor of a game's development that gamers care about. It's about flexibility, and extra performance is only one thing a developer could  choose to do with it.

7

Are you kidding me? Mantle help the PC gaming industry a whole much better. It got Microsoft and the Khronos Group to start updating their APIs. Before Mantle, the only viable API to use was DirectX (OpenGL not so much explained here in this thread: http://programmers.stackexchange.com/questions/60544/why-do-game-developers-prefer-windows) and even then there were problems like no viable way to make games for Linux/OSX and huge CPU overhead. This is what a monopoly does. No drive to innovate since you are the only one that has what everybody wants.

Mantle in the foreseeable future? I am not so sure. It could flop like Glide. However, I could say that if Mantle supports Linux and is open-source, it might have a bright future ahead of itself. But whether it truly succeeds or not, it brought the need of attention to the two current APIs developers have been crying for a rework.

With the new APIs coming out fixing the CPU overhead issue, the potential in graphical fidelity for games in the future will be exponentially high with the only limiting factor being the true power of the CPU itself since it is not bottlenecking the GPU.

8

As other have pointed out, Mantle is nothing if not good. You have to be on straight crack if you think the monopoly that Is DirectX is best for the games industry. Sure they have to use less different development resources because it is only one set of APIs but they ONLY work on Windows and xbox.

What about, Mac, Linux, PS3, PS4, Xbox 360, Xbox One, Wii, WiiU, Android, iOS, WinPho so on and so forth. While a very long stretch of the imagination it is conceivable that the open source nature of Mantle could mean one development for ALL platforms. 

This is clearly nonsense but it is what they are proposing and working toward. At least for Win, Linux and possibly Mac. One unified base that will work across all of them. How is that bad? Or at minimum worse than DriectX?

There is also the fact that while it will probably die, I hope it does not, it has given a very real shake and wake up to the current set. Direct X12 and Mantle are very inter-co operable for the initial information. And OpenGL is finally getting to a point where it can be taken seriously for gaming. 

There is also the very clear upside that Mantle would have the ability or rope in the likes of Open GL, open Al and open CL or what every form they take in the future. I for one do not want a future where things like G-Sync, Physx, and Cuda are splitting developers/gamers because they are no usable everywhere. Yes I am saying Mantle could be use on Nvidia hardware. AMD pointed this out them selves. This also offers the open source community some way to side step the quagmire that is Nvidia's refusal to work nicely with the open source community. 

Again another advantage is the bare metal access to hardware. As it stands you have to use Direct X to ask for access then pass through layers of APIs just to get to the hardware. Mantle works by stripping out as many interfering layers as possible and getting information where it needs to be faster.  

I would very much like to end this with, "Obvious troll is obvious, 4/10" but I get the feeling you are serious. 

 

I am neither a programmer nor an engineer so this information may not be exact.

9

I have a lot of experience programming in the same languages that games are developed in. I've been working in C++ for over 5+ years. I understand how APIs work. I, however, haven't had any significant time to look at Mantle.

Most of you have been very helpful in your posts. You have actually succeeded in changing my mind about Mantle. If Mantle moves DirectX and OpenGL towards lower overhead and faster performance, I am all for that. Hell, I want that to happen because it helps the industry that I am aiming to work in.

My concern was "fragmentation" as Pyrophosphate pointed out. His overall post regarding that was good and informative. I think that in conjunction with his reasoning, if OpenGL is forced to become better because of Mantle, it's win-win-win.

The only other concern was that the graphics API has a lot to do with the performance of a game. For the devs to split and program for both takes time out of just making one solid experience. If the two APIs act significantly different from one another it can affect how the program is actually designed. It could potentially increase development time. Another concern that I had was about NVIDIA hardware, because it is not clear if NVIDIA ever plans on supporting mantle. Though "Zibob" pointed out, that it is possible to rope in other APIs. Which is interesting.

I suppose those points are moot if DirectX and OpenGL get much better.

I very much so support the open source world. I love ArchLinux. The main drawback to Linux right now is the X system. Which is hopefully going to be replaced by Wayland soon. I hope that Mantle and OpenGL have reasonably low overhead on Wayland.

It however saddens me that the community was somewhat vicious to my post. I thought that it was implied that I am always willing to change my mind with the presentation of new evidence or opinions on that evidence that are compelling. This website is much better than the likes of YouTube, but it was disheartening to think people thought that I was trolling or that my question could be brushed off with a slight of hand. It is possibly my fault, because the internet is usually full of people set in their ways.

 

10

its a disruptive technology, so there will always be some negatives to it.

What made me laugh was that initially a lot of people were scratching their heads why bare metal api's were even necessary in the first place.

a lot of software houses wont bother as they will continue to do what they have always done and target dx9 for the widest possible adoption (cough ubisoft cough).

I have no great hopes for dx12, you just know that microsoft in their infinite wisdom are going to do a 'window 9 ONLY' shitty move which means no developer will even bother (see above reason).

If they release it for windows 7 up then dx 12 has a chance of killing mantle COMPLETELY dead right from the start and I will be over the moon (although if I was a betting man I would say in all likelyhood it will most likely be 8.1 upwards), if they do their usual shitty practice then mantle still serves a purpose.

The level of inefficiency in the software layers (specifically direct x but also the bloatware that is known as windows in general) is totally killing performance on tablets with win 8.1 installed on them.. and sadly that seems to be where microsoft is focusing their attention.

Most of us on this forum have got monster graphics cards and yet as soon as you slap windows on your machine they become nothing more than mediocre (or at the very least marginal gain).

it makes me cry

sorry for the rant, rant now over :D

11

Not meaning to be vicious, just kind of amazed. And I honestly half way through thought, "I have been had, and this is a troll post " turns out not. Which is nice. More information is always good. Sorry about that. 

 

EDIT: I have been out of the Linux loop for too long what is this Wayland you speak of? A replacement for X? Will it be easier to get along with?

12

I thought DX 12 was already confirmed for windows 7..

13

if this is true then it makes me very happy.

14

I'm sorry if you thought I acted vicious towards you. From your initial post, it sounded like you were very serious but I'm glad you are open to criticism and more information.

Got to remember we are one big happy family.

15

I'm just happy they're pushing the envelope and making people realize there are other areas to improve upon.