Mantle and Frame Latency

[H@mmer]

Does mantle come to a grinding halt in some situations on bf4 eyefinity? Yes.

Particularly if i'm in a tank and I engage the aps and a tankbuster is strafing me, the close up explosions can dive bomb the game to single digits, whereas on dx it'll stay around 35-40fps. Its a pretty common situation while driving a tank around and currently mantle doesn't do a good job of providing a good experience at 5760x1080.

I get the same at 1920x1080. If 2 jets collide and explode both my GPU and CPU graphs spike and my FPS come down. I don't know if they go in to single figures though.

 

[Gerard]

I get the same at 1920x1080. If 2 jets collide and explode both my GPU and CPU graphs spike and my FPS come down. I don't know if they go in to single figures though.

Yeah it does seem linked to large\close up explosions, pain in the ass as its pretty smooth upto that point then the fps just dive bomb. Should be investigated and fixed but apparently mantle still has a memory leak in b4 that's unfixed so im not holding my breath.

 

[tommybhoy]

BF4 using my 670 v Mantle 7950-no contest on MP tbph.

Mantle+freesync could be very interesting too regarding frame times.

A lot of ignorance shown where Mantle benefits, despite multiple users using high end gpu's and lesser powered Intel/AMD cpu's, this is often brushed under the carpet and is a major plus looking forward expecting DX12 to further implement negating the need for high end cpu's giving more lee way to budget for other tech/larger game libraries.

Mantle isn't without it's faults, but it is very early days and should only improve with further adoption-AMD funded or other wise.

Anyone have a game save profile for BF4? I still not completed it. I need a profile where I can select the mission I want to play.

Going to run some tests over the weekend.

I'll send mine over later tonight when I get a chance.

 

[GoogalyMoogaly]

BF4 using my 670 v Mantle 7950-no contest on MP tbph.

Mantle+freesync could be very interesting too regarding frame times.

A lot of ignorance shown where Mantle benefits, despite multiple users using high end gpu's and lesser powered Intel/AMD cpu's, this is often brushed under the carpet and is a major plus looking forward expecting DX12 to further implement negating the need for high end cpu's giving more lee way to budget for other tech/larger game libraries.

Mantle isn't without it's faults, but it is very early days and should only improve with further adoption-AMD funded or other wise.



I'll send mine over later tonight when I get a chance.

I thought my 670 did a better job than my 7950 in BF4, that was pre-Mantle though.

 

[tommybhoy]

Couldn't comment on a straight DX comparison, just DX v Mantle.

 

[MilkyJoe]

When will mantle be realised for crysis 3? i though i saw some one testing crysis 3 with mantle but i cant find anything about enabling it.

Also i see people buy latest graphics cards and cpus spending loads upgrading from a 780 sli to 980 sli, but it kinda annoying knowing that DX is holding back your performance when u spend like £1000 on gpu`s.

 

[tommybhoy]

iirc, there was fake pictures doing the rounds, 99.9% sure C3 won't get Mantle support, game support on older titles usually stops as they move on to other projects.

 

[layte]

All this proves is that Mantle provides better frame pacing in some circumstances when compared to AMD's DX driver in BF4. Now is that down to Mantle being good, or AMD's DX driver being bad as R&D resources are diverted from DX to Mantle?

 

[MilkyJoe]

i think dx on amd is a bit worst on some games but it also depends on driver u use, here is a link about mantle http://techreport.com/review/25683/delving-deeper-into-amd-mantle-api/3 i jumped a head abit but u can read from start if u like.

 

[tommybhoy]

All this proves is that Mantle provides better frame pacing in some circumstances when compared to AMD's DX driver in BF4. Now is that down to Mantle being good, or AMD's DX driver being bad as R&D resources are diverted from DX to Mantle?

Spoiler

You can't be worse when you are better than your competitor on DX never mind Mantle.

 

[layte]

I'm not sure if 6 month old tests are indicative of the current situation, and I'm not interested in bringing NV into this.

 

[flopper]

All this proves is that Mantle provides better frame pacing in some circumstances when compared to AMD's DX driver in BF4. Now is that down to Mantle being good, or AMD's DX driver being bad as R&D resources are diverted from DX to Mantle?

Some say Nvidia write so good drivers, why then are they worse than amd in regard to framepacing? This with the notion that nvidia had years to work the framepacing before amd catched up and passed them.

Mantle just icing on the cake.

 

[layte]

Now, both of you please take note, that at no point did I raise the spectre of Nvidia. I'm not interested in making this yet another tired and pointless mud slinging match.

Who is to say that AMD could have not made significant reductions in DX frame pacing spread if the resources put into the BF4 driver had been applied to the general DX driver. Such efforts would benefit more than a single bug ridden game.

 

[GoogalyMoogaly]

Spoiler

You can't be worse when you are better than your competitor on DX never mind Mantle.

I don't get it.
The only one there that shows Nvidia and AMD on the same graph is the middle one.
Rather than use 2 290X cards they've gone for a 295x2, which seems like a bit of cheat as it might have added hardware to help framepacing in Crossfire. It would also have been nice to have put that line infront of the SLI one so we could see it.
When you then compare the 1 card vs 1 card, they both jiggle about all over the place but the Nvidia one is lower. Is it a case of higher is better with frametimes?

The other thing, as with fps really, is it's one thing to be able to show it on a chart but another if someone actually feels it. Much like the AMD people were saying back when Nvidia had the better framepacing.

I think the Mantle charts clearly show Mantle is better, but at what point does it become noticeable? Does it vary from game-to-game?
Maybe we could see some graphs from games that weren't produced by AMD partners?

 

[humbug]

Your looking at charts created by an Nvidia invention to measure game smoothness ^^^^

Now, both of you please take note, that at no point did I raise the spectre of Nvidia. I'm not interested in making this yet another tired and pointless mud slinging match.

Who is to say that AMD could have not made significant reductions in DX frame pacing spread if the resources put into the BF4 driver had been applied to the general DX driver. Such efforts would benefit more than a single bug ridden game.

This is what you said.

All this proves is that Mantle provides better frame pacing in some circumstances when compared to AMD's DX driver in BF4. Now is that down to Mantle being good, or AMD's DX driver being bad as R&D resources are diverted from DX to Mantle?

Tommy posted a slide which shows AMD Frame Pacing performing better than Nvidia in DirectX.

The question cannot be answered unless you have a yard stick, the only yard stick available are comparisons to Nvidia, he answered your question, the answer is AMD do very well in DirectX, that is unless you consider AMD Bad and Nvidia even worse.

He's responding to you in your question, your response should be something along the lines of "ok thank you Tommy"

 

[almighty15]

All these Mantle threads and yet most reviews show that DirectX11 is still generally faster then Mantle on AMD cards.

 

[layte]

Your looking at charts created by an Nvidia invention to measure game smoothness ^^^^



This is what you said.



Tommy posted a slide which shows AMD Frame Pacing performing better than Nvidia in DirectX.

The question cannot be answered unless you have a yard stick, the only yard stick available are comparisons to Nvidia, he answered your question, the answer is AMD do very well in DirectX, that is unless you consider AMD Bad and Nvidia even worse.

He's responding to you in your question, your response should be something along the lines of "ok thank you Tommy"

He answered his own question by going off on a tangent about Nvidia, something both you and he usually moan about in the many Mantle and other AMD related threads we have.

I guess I can answer my own question. We cant know if that is the case, but what we do know is that AMD have a vested interest in making Mantle look as good as possible.

 

[shankly1985]

All these Mantle threads and yet most reviews show that DirectX11 is still generally faster then Mantle on AMD cards.

Well you not reading what this thread is aiming at are you. You just like the other people who posted about Frame rate being higher than Mantle etc

Little tip this is about how mantle is smoother than DX even when the frame rate is lower. Its Latency

What I would like is for some users of nvidia if you can to post a frame time of a single player mission.
Am going to record a play through of a mission to post on here and then we can follow it...

I going to run it on max setting on ONE GPU First on DX and then Mantle, with no Vsync or Frame rate CAP.

So what mission on the single player you think is best? Tommy send me his complete save file thanks mate.

 

[Ian Evey]

I agree with GoogalyMoogaly, the middle one shows a 780ti and a 290x and the 780ti is lower than the 290x and just as smooth. No mention of sli vs xfire in the op is there?

Not sure why someone posted this tbh. > Tommy posted a slide which shows AMD Frame Pacing performing better than Nvidia in DirectX.

I am sure these kind of threads are purposely created to fulfil some peoples self belief in that one brand of card is better than the other, knowing full well no matter how hard you try, its all irrelevant to some and fact to others and there will always be counter opinions.

Personal choice is all that matters.

 

[Gregster]

I don't wish to get tied up in this thread but want to point out that the slides Tommy posted (I believe) are from PCPer and in that, nVidia had a driver issue with SLI in BF4, which nVidia responded would be fixed.

Just clarifying and not getting involved