Does Mantle... work enough to rock your world and change your next Graphics Card?

[DanceswithUnix]

The shoddy drivers argument doesn't stand up - if AMD couldn't make good drivers then Mantle would suffer from the same problem - it's the same team for both as far as I can tell. However LtMatt's statement that it's delivering a superior experience aside from CPU efficiency and fps don't stand up either - there are no extra experience factors that Mantle enables outside of those so anything he's experiencing is a placebo effect.

Too lazy to try and find the graphs, but Nvidia drivers do actually scale down better to lower end CPUs than AMD cards. Not as dramatic as Mantle shows, but there is a difference.

Last card I bought was AMD and is mantle capable (the only mantle card in the house), but that coice was due to old prejudice of my wife against Nvidia drivers. I think that is unfounded, but as the 260X seemed the best card for the job anyway there wasn't any point in arguing for the Nvidia drivers

[Unstable Sloth]

i found mantle very underwhelming so id prefer shield game stream over mantle

[jigger]

I'm seeing some very nice performance gains from Mantle and more than enough to be a major factor when considering any upgrade. So IMO yes, Mantle was like getting a free upgrade on top of my upgrade.

[KeyboardDemon]

I'm seeing some very nice performance gains from Mantle and more than enough to be a major factor when considering any upgrade. So IMO yes, Mantle was like getting a free upgrade on top of my upgrade.

That's excellent, what sort of gains are you talking about?

[jigger]

Between 10-20% I would say.

[KeyboardDemon]

Between 10-20% I would say.

Is that across the board or with specific titles?

Also what hardware are you running? I do remember that when Mantle first came out the Beta 14.xx drivers were showing their biggest gains on the lower end of the range so the 270/270X cards were getting around 15% or 9 to 11 fps gains while the 290/290x were averaging around 1 to 1.5 fps gains in BF4, typically the 290's were going from 94fps to 95.5fps and the 270's went from 54 to 65.5 fps. Most of the friends I know that are using Mantle are saying that their frame rates are more consistent now and not really that much higher in games like BF4, while running the games in DX11 mode leads to bigger swings between the highest and lowest fps scores.

[jigger]

I've got a system with a pair of HD7950's and another with a pair of R9 290's. I've only got BF4 and Thief that make use of Mantle. I'm not sure if any of my other games support Mantle yet.

[KeyboardDemon]

I've got a system with a pair of HD7950's and another with a pair of R9 290's. I've only got BF4 and Thief that make use of Mantle. I'm not sure if any of my other games support Mantle yet.

I'd be interested in finding out what difference it makes to your frame rates if you test each game in DX11 mode and then again under Mantle, so do about 5 Fraps benchmark runs and take the average of each settings for each rig. In the first Mantle beta driver release it would have been the 7950's that realised the biggest performance gain. IIRC.

[directhex]

I'm not spending money on AMD.

My job is Linux based, I use Linux day-to-day, and now there's Steam on Linux, there's no excuse for the terrible TERRIBLE performance AMD offers me.

When AMD are serious, I'll consider them.

[KeyboardDemon]

I'm not spending money on AMD.

My job is Linux based, I use Linux day-to-day, and now there's Steam on Linux, there's no excuse for the terrible TERRIBLE performance AMD offers me.

When AMD are serious, I'll consider them.

You have a system with a GTX6600, this was launched in 2004, are you seriously trying to suggest that AMD have still yet to produce a card that can outperform it after 10 years of developments?

[Agent]

You have a system with a GTX6600, this was launched in 2004, are you seriously trying to suggest that AMD have still yet to produce a card that can outperform it after 10 years of developments?

He never said anything of the sort!

I suspect the GTX6600 is a typo, as there was no such card (only a 6600 GT).

What Directhex was saying, is that in answer to the thread title, he would not consider a AMD card as competitors offer a much better experience / performance under his OS.

Last time I used ATI on Linux is was truly horrendous. It was a few years ago now, but it really was unbearable.

edit - http://www.phoronix.com/scan.php?pag..._high514&num=5

An AMD 290 is being matched by a 760GTX here. The 290 does a bit better on some of the other tests, but in general, NV is a much better choice for Linux users:

hile our comparison was limited on the AMD side due to the lack of AMD's cooperation, overall, it seems the high-end NVIDIA GeForce graphics cards continue to smash the AMD Radeon graphics cards under Linux when it comes to performance with the proprietary drivers. The AMD Catalyst Linux driver has been improving since Valve's been pushing Linux gaming, but still the NVIDIA proprietary Linux graphics driver is delivering top-notch OpenGL performance and continues to be the leading recommendation by game developers and gamers alike.

http://www.phoronix.com/scan.php?pag...uxgaming&num=1

NV also seems to be the better choice at 4K on Linux:

Overall, for any Linux gamers wanting to pursue a 4K monitor right now, you need a rather powerful upper mid-range to high-end graphics card to run common open-source Linux games and even Valve's popular Source Engine titles. The GeForce GTX 780 Ti did the best while the Radeon R9 290 was running into a number of issues on the Catalyst 14.6 Beta driver at this resolution.

[KeyboardDemon]

@Agent: Thanks, when I read the post by directhex it came across as the hardware wasn't up to the task, even without looking at the links you posted I can see it is more a case of AMD not supporting Linux as well as nVidia or even sufficiently for a Linux user to get on with the tasks they want the card for and I can only assume that this also means that the Linux community as a whole have not found a way to create their own uber drivers to make the AMD cards work as well as the same cards running under a Windows OS.

[dfour]

I'd be interested in finding out what difference it makes to your frame rates if you test each game in DX11 mode and then again under Mantle, so do about 5 Fraps benchmark runs and take the average of each settings for each rig. In the first Mantle beta driver release it would have been the 7950's that realised the biggest performance gain. IIRC.

I did a batch of thief tests @1440p from low settings to ultra if your intersted I can post them for you.

[KeyboardDemon]

Yes please, that would be awesome.

[dfour]

No problem Im at work now but will post them later for you.

[chrisFletcher]

The one thing Mantle does seem to have done is to put the boot up Microsoft's behind in sorting out the next version of DirectX, it looks like it's attempting to improve performance in much the same way