• 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.

Mantle Feedback/Bugs

meldarthx said:
what's the best way to change the frame pacing? I got BF4 working again LOL
Click to expand...

Type this into the console RenderDevice.FramePacingMethod followed by the number.

2-stock
1-works well
0-disabled

I use RenderDevice.FramePacingMethod 1. If you want the extra fps run 0, but the hit from using 1 is quite small.

You can also place this command in your user.cfg file to save having to type it each and every time you play the game.
 
What's the point of Frame Pacing Method 2 if all it does is add extra stuttering and lower frame rates? Hope this is fixed in the next driver

LtMatt can you test it in thief?
 
Neil79 said:
What's the point of Frame Pacing Method 2 if all it does is add extra stuttering and lower frame rates? Hope this is fixed in the next driver

LtMatt can you test it in thief?
Click to expand...

It doesn't add extra stutter, it makes the game incredibly smooth. Those graphs paint a misleading picture at times. It does hurt fps though because the render times are ridiculously consistent.

Crossfire not working in Thief on Mantle yet. These various frame pacing methods are unique to BF4 currently.

The funny thing is, even a i7 2700k @3.9ghz with 1600mhz ram is not bottlenecking a 290 crossfire setup when using mantle.
 
Last edited:
Mines is running great @ 2 :D

Also i ran my 2700K at stock last night and didnt notice any difference in perf from 4.5GHz. Up to 30% CPU usage :) Thats at all High and 4xMSAA.

Oh.. welcome back Matt :)
 
Last edited:
MyBrains said:
Mines is running great @ 2 :D

Also i ran my 2700K at stock last night and didnt notice any difference in perf from 4.5GHz. Up to 30% CPU usage :) Thats at all High and 4xMSAA.

Oh.. welcome back Matt :)
Click to expand...

Same and cheers. :)

Neil79 said:
Did here and with that other user, without setting frame pacing to 1 I noticed it was giving the odd stutter, where as DX was smooth.

DX = Smooth
FP 1 = Higher frames, smooth
FP 2 = More frame rate drops and stuttering
Click to expand...

Correct, it did initially. However it works fine for me now so im wondering if perhaps it was fixed between then and now. There has been one game update and a server update since i last tried it. Still lowers my fps, but method 2 works fine (smoothly) for me now.
 
LtMatt said:
I would expect this type of game to get the most possible benefit from Mantle being cpu heavy.
Click to expand...

Surely Mantle frees up CPU from rendering graphics though? RTS games are often pretty simple graphics but hard on the CPU because of CPU tasks aren't they? I always assumed performance was an issue due to the amount of things happening in parallel as opposed to the complexity of what was being displayed.
Total War games are probably an exception there.
 
GoogalyMoogaly said:
Surely Mantle frees up CPU from rendering graphics though? RTS games are often pretty simple graphics but hard on the CPU because of CPU tasks aren't they? I always assumed performance was an issue due to the amount of things happening in parallel as opposed to the complexity of what was being displayed.
Total War games are probably an exception there.
Click to expand...

Mantle is so efficient on the cpu side of things id expect big gains. Time will tell.
 
GoogalyMoogaly said:
Surely Mantle frees up CPU from rendering graphics though? RTS games are often pretty simple graphics but hard on the CPU because of CPU tasks aren't they? I always assumed performance was an issue due to the amount of things happening in parallel as opposed to the complexity of what was being displayed.
Total War games are probably an exception there.
Click to expand...

Mantle does save some (pure) CPU overhead but one of the biggest areas it gives gains in is actually "API" bottleneck rather than CPU bottleneck but as its not in a place where the GPU is busy its referred to as a CPU bottleneck - while this is terribly over simplified:

Mantle:

CPU/GPU handshake
Draw 1000 boxes
CPU/GPU handshake

DX:

CPU/GPU handshake
Draw 1 box
CPU/GPU handshake
CPU/GPU handshake
Draw 1 box
CPU/GPU handshake
CPU/GPU handshake
Draw 1 box
CPU/GPU handshake
CPU/GPU handshake
Draw 1 box
CPU/GPU handshake
CPU/GPU handshake
Draw 1 box
CPU/GPU handshake
...repeat for all 1000 instances

In the DX example a lot of the time both CPU and GPU are idle waiting for the API to finish doing its stuff (rather than a traditional CPU bottleneck where the CPU is too busy to respond to rendering stuff) the change does increase effectively utilisation of the CPU however.

(This is a hugely over simplified example)
 
Last edited:
Good read matt, after reading, it looks like the 7 series high vram usage is needing fixed by Dice on the engine side, as it's under the developers control.

Amd driving api's forward, looking forward to DX 12 (DXMantle:p) so that Nvidia users get a slice of the pie it's too good to keep to ourselves.
 
You must log in or register to reply here.
Back
Top Bottom