• 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.

Upcoming games to support Mantle

Look bottom left and the performance graph. lol Go record some gameplay yourself and post back your Graph.

No different to the graphs humbug has already shown, doesn't mean a thing because as said above you can't perceive the difference between a few 0.1 seconds, actually wait you'll probably tell me you can.
 
Where did you get 'a few 0.1 seconds' from? its nothing to do with that, its about dropping frames and consistency ^^^ :)

I don't actually know these people and was just something I was asking in game. Good of you to help but I am not sure I will see them again.

Now then, as for Mantle's memory leak, care to divulge on how to fix it? I can maybe pass that info on when in game next time :)

your welcome.

I edited my original response to you to answer Roff who had the same question. its not a Memory leak, i though i told you that once or twice already?
 
Last edited:
No different to the graphs humbug has already shown, doesn't mean a thing because as said above you can't perceive the difference between a few 0.1 seconds, actually wait you'll probably tell me you can.

You even reading what i saying? The smoother frame rate for me is when big changes happen. Note in the video I be running 180fps and then drop to say 150fps that DROP on DirectX gives me a slow down effect and I feel it. On Mantle I DONT get this and it Continues to feel smooth.

GET ME?
Maybe am just very sensitive to frame rate change?
 
Isn't the memory issue 7000 series only?

I don't properly know how to answer this. But that can't possible be the case unless it's AMD at fault. 7XXX a lot being rebranded to RX. Take the 270X, if that's hunky dory on Mantle, and the 7870 isn't, then it's AMD's drivers at fault (And we've been told this isn't the case)
 
I don't properly know how to answer this. But that can't possible be the case unless it's AMD at fault. 7XXX a lot being rebranded to RX. Take the 270X, if that's hunky dory on Mantle, and the 7870 isn't, then it's AMD's drivers at fault (And we've been told this isn't the case)


Mantle is hunky dory on the 7870, it just doesn't have enough buffer to run the Frost Bite 3 engine with mazed out textures in Mantle. turn that down and its perfect. i know, i had one.
 
I don't properly know how to answer this. But that can't possible be the case unless it's AMD at fault. 7XXX a lot being rebranded to RX. Take the 270X, if that's hunky dory on Mantle, and the 7870 isn't, then it's AMD's drivers at fault (And we've been told this isn't the case)

Well seeing as Mantle works on 7000 series for all other games that support it and only BF4 that have the issue. And the fact Mantle is done by the Devs and not by Driver you would think Dice are to blame for the memory issue on that game.
 
Mantle is hunky dory on the 7870, it just doesn't have enough buffer to run the Frost Bite 3 engine with mazed out textures.turn that does and its perfect. i know, i had one.

The context was to Frosbite running Mantle.

And Thief using more VRAM in Mantle than DX, so potentially, it can have a problem in other games as it's still less efficient, even if Thief's running perfect.
 
Well seeing as Mantle works on 7000 series for all other games that support it and only BF4 that have the issue. And the fact Mantle is done by the Devs and not by Driver you would think Dice are to blame for the memory issue on that game.

Don't all current frostbite mantle games have problems? (Matt's said they do......)
You actually haven't really challenged my point either.
 
This shows the smoother frame rate.
BF4_1920x1080_PLOT_0.png

BF4_2560x1440_PLOT_0.png


http://www.pcper.com/reviews/Graphi...e-CrossFire-Early-Performance-FCAT/High-End-C
 
Another pointless graph chucked in there that's already been posted, just goes to show how much attention you pay to anything shanklol
 
There's a guy on the 2nd page of this very thread said he notices no difference between DX11 and mantle as well.

Personally I don't actually understand how people are noticing 0.01 second frame time differences when playing a first person shooter at 60+ FPS. Would people even have noticed this without charts like the ones above?

Not a chance, it's placebo effect, they're seeing what they want to see.

It's really very noticeable to most people, but a lot of people are in denial because real world testing and results often show things that are inconvenient to their prejudices.

I can give you some examples here using Nvidia's own FCAT tool which measures frame latency, variance as well as things like dropped frames. We will compare AMD's Mantle vs Nvidia's DX11 implementation in Battlefield 4. Battlefield 4 is a good title as it has been out for ages and is well optimised for both sides and it saw a nice gain from the Nvidia wonder drivers.

AMD/Nvidia DX11

xIJTTiV.png


sF93km1.png


Variance and spikes will represent itself as hitching or stutters during gameplay. It's very easy to see and feel these while playing because typically the game is smooth for the most part.

Looking at the above chat we can see that AMD is offering a smoother experience at 1440P under DX11. Much less fewer spikes and hitching and a tight consistency of frame times. If we look at the frame variance in MS, Nvidia are offering 7+ in variance whilst the AMD solution offers 3+. Nvidia are offering more than double the variance in their frame times using DX11, ouch! Single cards are offering 1 for the 290X and just over 1 for the 780TI. Both very good for single cards, no issue here.

Lets have a look at Mantle.

MNU6Ajt.png


QxVFKIP.png



The Mantle line is completely flat. There is not one single spike, not one instance of variance, microstutter, hitching, nothing, perfection. Surely it can't be better than single card, can it? Lets look at the variance chart in MS. Again the crossfire setup on DX11 offers a ball hair over 3MS variance. Mantle however offers 0.5MS of variance! That is lower than a single card can manage under DX.

That is why the majority of us rant and rave about Mantle. It's head and shoulders (woo shampoo pun) above anything else out there.

Remember, this is using Nvidia's own tool which they created to highlight the problems AMD used to face with frame pacing, variance and dropped frames.
 
Last edited:
Yeah but if that is Nvidia's own tool, how can we trust it? I thought that "Nvidia are evil and lie about everything whilst trying to gimp AMD performance"... :D
 
The context was to Frosbite running Mantle.

And Thief using more VRAM in Mantle than DX, so potentially, it can have a problem in other games as it's still less efficient, even if Thief's running perfect.

what you have to remember is Mantle has the option of pre-loading textures so you don't get texture filling and slowdowns during game play, thats a good thing but will use more vram

Its only BF4, possibly BFH too, i don't know, if so its a problem with the engine, not just DICE, i suspect it is the engine. BF4 has had its fair share of memory management issues in DX, which have been fixed. they just haven't been fixed for mantle, possibly because the pre loading feature is busted in that, it never unloads the textures from the last round, it just builds up over time, only in BF4 does it do that.
 
Back
Top Bottom