• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

When is the AMD roadmap out?

I would be interested in knowing as well, since it may influence my proposed build. I have been looking around the web for the last few days and found nothing thus far.
 
Am glad we finally get to see some AMD action, hopefully all mantle questions are answered.

Then everyone can go consider options or simply Core i7-4770k :)
 
Hope for some cpu news soon else will just go 4770k and be done with it

Agreed unless Mantle going to give 20-30+fps more or bring a level of stability and consistency or something new to the table, 4770k is the best option. Heck they should give extra 20fps via the AMD cpu alone and then 20 fps via the gpu !

10 or 20fps won't be worth the effort or energy and wont even catch up even to an old core i7 over-clocked to 4.2ghz+.

I think we already could answer 4770k the better option anyhow just nice to see if it was going to really bring any difference :D
 
Is this Mantle thing going to be available for linux and OSX?

Is it a replacement for Direct X like OpenGL?

If it doesnt support linux and OSX I see3 it will fail. Why would developers want to use it instead of OpenGL or DX?
 
Regarding Mantle, tomorrow we will know a the developers summit. Just patience.


@HighlandeR
As for upgrade & roadmaps, if you have an 2500K overclocked just hold until Spring.
Intel 1150 is dead so is 2011. There are going to be no new CPUs on both lines. The new socket is 2011-3 and not compatible. (with DDR4 probably)

AMD weknow Kaveri APU are FM2+. We do not know about Steamroller if comes to AM3+ (if at all), and we know that FX is dead. After that official roadmap will be announced December with start of sales in February.
 
Agreed unless Mantle going to give 20-30+fps more or bring a level of stability and consistency or something new to the table, 4770k is the best option. Heck they should give extra 20fps via the AMD cpu alone and then 20 fps via the gpu !

10 or 20fps won't be worth the effort or energy and wont even catch up even to an old core i7 over-clocked to 4.2ghz+.

I think we already could answer 4770k the better option anyhow just nice to see if it was going to really bring any difference :D

So in your mind Mantle is fail if an AMD rig (CPU and GPU), that currently runs a game (e.g. BF4) at 90-100 fps average on ultra settings at 1080p, can't then reach 130-140fps?

Seriously? Please get real. 10-20fps can be a massive increase and quite a bit more than a 10-20% improvement in fps. This is especially so in demanding games on ultra at 1080p+.

Telling us that you don't think a 10-20% increase in performance is worthwhile, when AMD owners with the correct hardware will simply have to do nothing (i.e. no extra hardware and no cost) to obtain this boost, is just naive. People upgrade their hardware for a lot less (e.g. Sandy to Ivy and Ivy to Haswell springs to mind).

Of course we don't know what improvements Mantle will actually bring yet but reading this sort of presumptuous comment isn't particularly helpful to others.
 
Last edited:
Who knows what's happening with it, it has new CPU's coming to it : refresh, who knows what that will entail.

I don't know what's happening with broadwell, all I know is there's always scare mongering, I can't see my launch day 1150 board taking a Broadwell.
 
So in your mind Mantle is fail if an AMD rig (CPU and GPU), that currently runs a game (e.g. BF4) at 90-100 fps average on ultra settings at 1080p, can't then reach 130-140fps?

Seriously? Please get real. 10-20fps can be a massive increase and quite a bit more than a 10-20% improvement in fps. This is especially so in demanding games on ultra at 1080p+.

Telling us that you don't think a 10-20% increase in performance is worthwhile, when AMD owners with the correct hardware will simply have to do nothing (i.e. no extra hardware and no cost) to obtain this boost, is just naive. People upgrade their hardware for a lot less (e.g. Sandy to Ivy and Ivy to Haswell springs to mind).

Of course we don't know what improvements Mantle will actually bring yet but reading this sort of presumptuous comment isn't particularly helpful to others.

Yes that is what I believe in and I have seen it before with AMD 3dnow quake patch from 15 years ago.

I would love to see 20fps+ or cpu cores and gpus being rendered far better something that really brings a difference, otherwise what was the point ?

If its not going to bring anything substantial to the table or get adopted and supported, it will end up just like it did with 3dnow.
 
Mantle could and Hopefully be the game changer but it will never make up for the fact that the 4770k is simply the better chip amd need to bring something to rival it
 
AMD don't need to even reach the same speed (though it would be nice) they just need to get near and bring it in at a price that takes all the i5 and some of the i7 sales. The 8320's have shown the potential that steamroller could bring, but given how quiet they are I wouldn't recommend holding off if an i7 is in budget.

Also referring to Haswell as a dead end, who cares, a clocked i7 4770k is going to run anything you throw at it. By the time you update that you would be updating your whole system.

edit: plus mantle will benefit intel users as well, its only the AMD gpu thats required isn't it.
 
As much as I would love to see Steamroller/Excavator on AM3+ socket motherboards, it is no great loss if they are only on FM2+ socket motherboards, so long as the CPU side of performance is there. This is the general direction that the industry is moving, AMD started the APU phenomenon, and just look now--practically all of Intel's consumer processors are in fact APUs.

It is ironic, but AMD needs to play catch up, and to do that I think they need to focus--as supporting two different types of socket (and chip) at the same time is no easy feat. AMD have a clear advantage in the graphics department, they just need to get the CPU portion closer to what Intel has managed, and then they are back in the zone. However, it should also be noted that the industry is also moving towards tech such as OpenCL, which when used definitely makes up for the lesser CPU grunt.

About Mantle, as I understand it, it is intended to be an alternative to DirectX and OpenGL. Not a replacement, but an option, much as Glide was for those who remember that. What it actually does, is allow more direct access to the hardware than either DirectX and OpenGL, which allows developers much more control and removes a lot of the load from the CPU.

The boost in performance from doing that, will vary, because it depends on developers to put in effort. I mean it doesn't matter how good an API is, if developers are not trying to use that API (or indeed available hardware capabilities) to their full potential, we see that all too often already. The potential for increased performance is certainly nothing to turn your nose up at though, you can be sure of that. As for platform support, we know it is going to be on Windows, and I think it would be incredibly silly if Mantle does not come to Linux too. I haven't seen anything to suggest that it won't support Linux or OSX, so maybe we will find out sometime during these AMD conferences that are going on.
 
Last edited:
Is this Mantle thing going to be available for linux and OSX?

Is it a replacement for Direct X like OpenGL?

If it doesnt support linux and OSX I see3 it will fail. Why would developers want to use it instead of OpenGL or DX?

It requires a thin (lightweight) additional driver. Assuming they write one for other OSes it will be available.

It's not supposed to replace other APIs, merely allow top-flight studios to push the limits.
 
We do not know about Steamroller if comes to AM3+ (if at all), and we know that FX is dead. After that official roadmap will be announced December with start of sales in February.

There's very little chance of SR on AM3+. PD will survive through 2014.

In 2015 we'll either see a replacement, or AMD will release Carrizo (Kaveri replacement) in 3 and 4 module versions.

It makes little sense for AMD to waste resources on an SR refresh of FX, when Kaveri will sell like hotcakes. They're better off concentrating on a real replacement, whether that be part of Carrizo, or something else. Introducing SR FX next year would just postpone the former.
 
Back
Top Bottom