You guys remember Mantle, right? It was AMD’s effort to inject life into what was once a languishing area of improvement: 3D graphics APIs, the software that powers modern, real-time 3D. However, it appears the company is shelving its creation a mere two years after its inception.
In his review of the RADEON R9 Fury X, Anandtech’s Ryan Smith approached AMD about the state of Mantle and received an interesting response:
The situation then is that in discussing the performance results of the R9 Fury X with Mantle, AMD has confirmed that while they are not outright dropping Mantle support, they have ceased all further Mantle optimization. Of particular note, the Mantle driver has not been optimized at all for GCN 1.2, which includes not just R9 Fury X, but R9 285, R9 380, and the Carrizo APU as well.
The article mentions that a lot of what made Mantle sing has been adopted by Vulkan, the Khronos Group’s successor (of sorts) to OpenGL and competitor to Microsoft’s DirectX 12. Smith also reports that AMD itself has “thrown all of [its] weight behind Vulkan and DirectX 12”, going so far as to dissuade studios from using Mantle over these options.
This was pretty much the only way things were going to go once Microsoft lifted the lid on DirectX 12. I doubt Redmond has been sitting on its laurels, but Mantle likely put a bit of fire under its seat.
With Vulkan inheriting the work AMD put into Mantle, there is a silver lining and the API will live on, even if its original incarnation eventually goes the way of 3dfx’s Glide.
The AMD Radeon R9 Fury X Review: Aiming For the Top [Anandtech]
Comments
11 responses to “Mantle, AMD’s Next Generation Graphics API, Is All But Dead”
Mantle was a good kick in the pants for DirectX 12, but its philosophy was flawed such that many developers didn’t really feel it was worth the investment of difficulty. I was critical of the API when it was first released because of its heavy focus on performance at the cost of abstraction, and how that would make developers’ lives harder in the long run.
Hopefully Vulkan and its OpenGL-fluent development team will improve the abstraction issues Mantle had, and I’m looking forward to getting a look at the API when it’s released. Competition is usually good for consumers and OpenGL hasn’t had much pressure on DirectX for a long time. Here’s hoping Vulkan will make for a good competitive, high-performing and importantly developer-friendly API.
Defiantly helped get DirectX firing again. and agree with what your saying but also feel it was Nvidia’s 0 care in using the tech that was also a pretty big factor, no matter how much steam it picked up Nvidia were never going to use it. The attitude of Nvidia is pretty well known towards things of this nature, they have to be in control and everything is certified, sli, gsync etc that’s just how they work. AMD however have a more open and sharing attitude which has some pretty big benefits, even though they were willing to share mantle it didn’t matter.
So when a dev is looking at the technology they are going to use, while Mantle did offer some benefits the con’s of having to run directX anyway and Nvidia users not being able to use it added to its chance of failure. Then DirectX 12 starting moving and really cut down on overheads etc leaving Mantle with just mostly cons when considering using it.
Thanks for the update, was wondering what happened to Mantle actually. I regretted getting a new Nvidia card when I heard about this, now I just regret getting a 2gb 😛
Mantle was a test bed for some nice changes for graphics API, personally I hope Vulkan wins the day because its open platform where DX12 is LOCKED into Windows10 which is VERY VERY BAD!
Bad if you want to use anything other than Windows. But if we’re just talking about the Windows ecosystem, I’m all for DX12 being 10 Exclusive. There’s no reason not to upgrade.
According to Steam’s hardware survey, only about 4% of users are on XP or Vista, so 96% or so should qualify for a free upgrade to Windows 10 when it releases. As you say, there’s no reason not to upgrade.
It’s not just that, Windows 8.1 and up adds lots of performance improvements in all areas, 10 adds even more to that and then DX12 makes it perfect.
I agree, I’m not one of the anti-Windows 8 folks. I upgraded the week it came out and have loved it ever since. It’s a clear improvement in performance over 7. Here’s hoping 10 continues that trend, it’s hard to tell right now in the tech preview since it’s unoptimised.
lol saw this coming the second they announced it. The only question I have is how are AMD going to blame Nvidia for this cock up?
Logan, the title of your piece does not reflect the body and seems designed to just say something negative about AMD. In the body of your story you make clear that Mantle has become Vulkan so why not lead with that?
Kolreth statements like yours are hurtful and ridiculous. If we had not given up the code just as we promised we would be criticized for that, now we do and you call it a cock up?
If you want to live in a world owned entirely by our competitor have you given any thought to what that would look like?
That was not the intention at all Roy and I’m sorry you feel that way.
As you yourself write, the article makes it clear what Mantle’s future is — which is not as Mantle. At no point do I state the API was a “cock-up”. If anything, I highlight the benefits of AMD’s efforts to provide a competitive API and even if Mantle won’t go on as Mantle, it will as Vulkan.
As a games developer myself who has released a cross-platform title, I certainly don’t want a world ruled by NVIDIA. Anything that makes developing easier for Linux and Mac, count me in.
What is worth keeping in mind is that AMD can release Mantle at any time to run on Windows 7 and 8 platforms.
If the Windows 10 adoption begins to falter then AMD can certainly provide a better API to those who just do not wish to upgrade legacy hardware to Windows 10. Microsoft would have to respond with a DX12 patch for older Windows versions.
They will do that next year anyway.