• 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.

The Gamers Guide To Mantle

Unfortunately i did this video before AMD released the new drivers, but nonetheless i captured a bit of Mantle footage of Plants Vs Zombies. Didn't venture online, just had a quick bash on the single player challenge. :)

Watch in 1440P Quality, more to follow. :cool:


 
Is there a requirement tied to Mantle like 4xMSAA and Vsync is for DirectX? I was thinking this last night because a few titles annoyingly still ship without support for Ultrawide,120fps,8xMSAA,UHD etc. It annoyed me they never forced them into full on support in DX 11.2 or 12.


If Mantle did it would actually be a damm good step to make people put in effort or lose DX/Mantle support.
 
LtMatt said:
Mantle working well, scaling is good. Fps around 140-190. However whenever i fire a projectile with a plant there is a spike in the overlay which lowers fps. Not sure if it is a game or API issue, i suspect game. Guess i can't complain too much as im playing it before release. Fixed by using vsync as the fps drop is unoticeable then. Will put up some video footage and you'll see what i mean.
Click to expand...

So if it is an AMD tied in game then it must be the game's fault, but if its an nvidia tied in game then it must be nvidia

we've got your number ;)
 
andybird123 said:
So if it is an AMD tied in game then it must be the game's fault, but if its an nvidia tied in game then it must be nvidia

we've got your number ;)
Click to expand...

Confirmed it's a game issue, as it happens on DX as well. Nonetheless, Mantle performance is great and no driver update needed either, it just works out of the box. :)
 
Rofflay said:
Elaborate please? Frametimes? Min fps placebo?
Click to expand...

Mantle's frame times (the amount of time it takes a gpu to render each frame) are so smooth and consistent compared to what DirectX offers (for either vendor) that each game using it ends up feeling much smoother because of the consistency.

It's hard to explain, but pcper did some mantle testing a few months back, whilst Mantle was in it's infancy. They used the frame pacing method 2 during their testing, which was the stock method at the time. This method sacrifices raw average fps (15-30+ fps) in order to give incredibly tight and consistent frame times, something which no other API is able to get anywhere near. See this graph.

mbNpHa4.png

Look at Mantle, in the Orange.

pcper said:
Multi-GPU results are...different. The orange line representing dual R9 290X cards with Mantle on the Observed FPS graph is oddly flat and barely faster than the single GPU. The CrossFire D3D11 results are more or less what we have come to expect - 78% faster than the single card. A quick glance at the Frame Times graph shows the orange line of Mantle multi-GPU is nearly perfectly flat - the frame times are incredibly consistent even though they are not as fast as we'd expected.

At 2560x1440 things change a bit. Mantle CrossFire results are again very flat, but at least scale up within range of the DirectX 11 CrossFire scaling. Going from a single R9 290X to a pair with the Mantle API results in 56% average frame rate scaling. That is less than the performance advantage provided by CrossFire in DirectX but if you look at the Frame Times graph, you will again see that the Mantle implementation is incredibly smooth and consistent!

Mantle multi-GPU scaling is much better and provides more tangible benefits with this platform at 2560x1440 than it does at 1920x1080, but the "flatness" of the frame times is a bit confusing. It almost appears that the Frostbite engine's implementation of frame pacing puts so much weight behind having a consistent frame rate that it is willing to sacrifice peak frame rates. In the case of our higher resolution results here, I think that trade off is worth making. The lack of ANY hitches or frame time spikes provides a spectacularly smooth gaming experience.
Click to expand...

I spoke with Ryan after publishing these results and informed him that if you changed the frame pacing method with the games console you would get similar frame consistency but a big fps boost, easily surpassing the best DX could offer for either vendor. Unfortunately he was unwilling to mention this in the article or alter it to inform the masses. Shame as it would've been useful information to get out there. Use frame pacing method 1 to unlock the raw average fps power of Mantle, as i've posted about many times here.
 
Last edited:
Thanks i suspected it might of been that. Im wondering if it would make BF4 playable at 60fps as i dont quite feel it is for me. And i bet it is the frametimes doing it making it seem like 30fps for a millisecond. If i notice a drop from 60-57 which i do then im going to notice frametime delays too.


Im tempted to get my hands on one to see for myself. But it wont effect the games that annoy me apart from BF. Chivalry is the one i suspect has the most disgusting frame times known to man though. It would be a prime candidate for it as i have to hide it with FXAA and blur not to get motion sick.
 
LtMatt said:
Unfortunately i did this video before AMD released the new drivers, but nonetheless i captured a bit of Mantle footage of Plants Vs Zombies. Didn't venture online, just had a quick bash on the single player challenge. :)

Watch in 1440P Quality, more to follow. :cool:


Click to expand...

Do you have a raw file for that as YT is limited to 30fps.

And that game is looking interesting.
 
You must log in or register to reply here.
Back
Top Bottom