mantle

[xMishax]

Would like to ask how potent is AMD's Mantle technology? Havent read any news about it for a while so wanted to ask you guys.. Maybe anyone knows somethings?

[Guest]

Not worth it.

I tried it with BF4 and the performance was worse with Mantle.

[ShadowCaptain]

Still a thing but DX12 is meant to be a low level API, which pretty much renders mantle useless

[Interfectorem]

It has great potential, but personally, I've had mixed experiences with it. In Thief, Mantle worked pretty well, increased my FPS by, I think ~5 FPS. But in BF4, Mantle gives me some major stuttering, so I switch it back to DirectX. It will probably improve in the future with more and more games coming out with support for Mantle.

 

EDIT: I also forgot that the upcoming DirectX12 will be a "Mantle Killer", like @ShadowCaptain said. 

[Notree]

Ive tried all the mantle games. While I gained a few fps here and there it was anything special. Not a feature worth considering unless your using certain hardware.

[Lord Pantaloons]

It mainly shows improvement with low end GPU's like R7 250/250x/260/260x/265 while only showing minimal improvement with mid to high end GPU's.

[ShadowCaptain]

Honestly the real place mantle works is in a CPU bottlenecked system, then it has a much more profound effect

[typographie]

In general, the performance improvements are more and more tangible the weaker your CPU is. But its got its own set of glitches still, and ironically support is poorer on lower-end hardware. I've heard BF4 has (had?) a glitch where Mantle could cause serious performance issues if you're not using at least a 3 GB card.

[Manny Calavera]

It mainly shows improvement with low end GPU's like R7 250/250x/260/260x/265 while only showing minimal improvement with mid to high end GPU's.

mantle is more of a cpu overhead removal, so cpus like athlon perform better

[typographie]

mantle is more of a cpu overhead removal, so cpus like athlon perform better

 

It seems to improve performance all the way up to the FX-8xxx's, sometimes allowing them to become GPU-bound and close the gap some with i5's and i7's at times. You see very little performance improvement, sometimes even a decrease, with a high-end Intel processor, though.

 

Tom's Hardware posted an awesome, detailed performance review of Mantle a few months ago: http://www.tomshardware.com/reviews/amd-mantle-performance-benchmark,3860.html

[xAcid9]

It mainly shows improvement with low end GPU's like R7 250/250x/260/260x/265 while only showing minimal improvement with mid to high end GPU's.

It should be the other way around since higher end GPU usually bottleneck by lower end CPU.

 

2GB card still have problem with Mantle if u're using high res texture.

[D2ultima]

Mantle reduces the load on the CPU by giving more direct access to the GPU.

 

This means games are less dependent on the CPU to pre-render soft frames before feeding them to a GPU. This benefits a strong GPU paired with a weak CPU the most. If you use mantle on a i7-5960X @ 4.5GHz paired with a single 280X for example, it probably won't do anything. If you use it on a R9 295X2 paired with a A8 APU for example... you'll probably add 50fps or more.

[A10-FTW]

this looks good  :rolleyes:

 

[UnbendingNose]

Depends on your CPU, people with Intel quad cores barely see any performance gains because they're fast enough without mantle. I have an fx8320 and I get about 15 to 20fps more using mantle in bf4 because the fx8320 bottlenecks my 290 bad on dx11

[UnbendingNose]

It mainly shows improvement with low end GPU's like R7 250/250x/260/260x/265 while only showing minimal improvement with mid to high end GPU's.

False, its the other way around. Mantle provides less overhead on the CPU and direct access to the GPU. So systems with strong gpu's and weak CPU's benefit the most. My r9 290 and fx8320 gets 15 to 20fps more using mantle over dx11 because my 8320 isn't a great CPU for the 290.