• 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 Benchmark Thread

Hi LtMatt,

Hope the following makes up for some school boys over the last day or so, have a double run on mantle and DX11 for ya!

Just have a few thoughts following many many runs on this & I'm not sure how mantle and AfterBurner work together but with crossfire on DX11 both cards sort of move around together on % usage , however with mantle only the first gpu seems to have any major movement as the second gpu almost locks at 100% only dipping during a lag spike which seem to be random, seems like a totally different approach to crossfire & it seems to be helping.

Also frame times (that FCAT stuff) seem much better, as the game just feels & runs better in mantle. When i went from crossfire 6970's to crossfire 290's, the frame times where clearly better in DX on BF4 however its like the same jump again with mantle.

Anyway onto the results both DX & mantle done at the same settings

i5 3570k @ 4.2
Ultra preset 4xaa @ 1080p motion blur off FOV 90
Crossfire Tri-x 290's (stock 1000)

Mantle
ll56.jpg


DX11
4lcj.jpg


So that's up to 49% improvement for me,
 
uksoldierboy said:
Lol sounds like a plan to me bud!, you get rid of that green crap, name a pub and i'll join you for said bender where the only discussion will be Liverpool:D

In all seriousness though, i won't defend Mantle, if i think something is not quite right, i'll say, i don't see the point in blind loyalty to company that don't even know i exist.

AMD have done me no favours, we aren't friends, the only reason i have an AMD card in my system is that for me, at this moment in time, it offers me everything that i want at a price i'm comfortable paying.

Greg if you like i could benchmark while playing alongside you? i really do wish they'd do a benchmarking tool for this though.
Click to expand...

That is the biggest problem Recce. No damned, in built bench test. It would make life so much easier. I wasn't doubting your results but you can see the differences from Dx to Dx and from a neutral, they look massively different. My style of play (Rambo) brings roughly the same results time and again, so my play syle is pretty consistent, even if it is crap :D

We have BF4 Mantle so far, which for the most part looks promising and I take nothing away from that. I have been reading around at various sites and apart from the early teething problems, results are looking promising with 10% in some cases and over 100% in others (I do question that though :D).

Anyways, I will be getting a ticking off from Matt soon for being in here :p
 
Nice one Jock, will start adding those scores to the OP. That's a nice improvement. I think you were always in line for a nice bump with Mantle because of your i5. Have you tried disabling ULPS btw? That may get you some extra performance and even out that gpu usage a bit.

Gregster said:
Anyways, I will be getting a ticking off from Matt soon for being in here :p
Click to expand...

As long as you're not derailing the thread/trolling. Had enough of that already. :p
 
LtMatt said:
Nice one Jock, will start adding those scores to the OP. That's a nice improvement. I think you were always in line for a nice bump with Mantle because of your i5. Have you tried disabling ULPS btw? That may get you some extra performance and even out that gpu usage a bit.
Click to expand...

Yeah ULPS is off, for both DX & Mantle. Been keeping an eye on "AB" as i have 3 screens, so the "AB" app on another monitor whilst running with mantle, thats how i spotted the different style.

when UPLS was on the spikes happened more with mantle, also having the display overlay off seems to have reduced the spikes even more, i still get them but they are more random now & Mantle seems less likely to crash
 
Bit of an odd question here, considering I have a 7970 and a 7950. Do you know if it's possible for me to turn off crossfire and use my 7950 as my primary card? ... ya know just for more varied benches.

If not, no worries. I'll just see what kind of gains I'll have with just my 7970.
 
Just done a couple runs comparing dx11/mantle with a single 7970 and i5 2500k @ 4.8. Pretty much exactly the same as each other.

frametimelog009 is DX11 and 011 is Mantle :)

gg2Wmhp.png

I'm not entirely sure where those minimums are coming from, at no point am I feeling the game drop frames that much.
 
can someone bench bf4 with 5760x1080 and at 200% render scale? ultra settings of course, wanna see how it compares to my setup, those are the two resolutions i play it at
 
sockky said:
Just done a couple runs comparing dx11/mantle with a single 7970 and i5 2500k @ 4.8. Pretty much exactly the same as each other.

frametimelog009 is DX11 and 011 is Mantle :)

gg2Wmhp.png

I'm not entirely sure where those minimums are coming from, at no point am I feeling the game drop frames that much.
Click to expand...

Hi,

Not to throw a spanner in & after a few of my school boy errors i need to watch the wrath of LtMatt...however I need to ask, did you change the API then quit the game then boot back in before testing the new API?

The reason i ask, if you look at the mantle results there are always way more frames than the DX runs (about 10,000 +) with DX there are always around 7000, which both your results have?
 
Last edited:
sockky said:
Bit of an odd question here, considering I have a 7970 and a 7950. Do you know if it's possible for me to turn off crossfire and use my 7950 as my primary card? ... ya know just for more varied benches.
Click to expand...

LtMatt said:
If you physically swap the cards round, then yes.
Click to expand...

You don't need to swap cards, disable crossfire then just swap your monitor cable into the gpu you want to use.

You're slacking Matt.:p
 
Yeah, i'm 100% sure I did. But now you've made me doubt myself.

Ha ><. I'll retest if you think I really should.

edit. Just done another quick go, trying to follow as closely as I did with the last run and frames were almost identical. Who knows eh.
 
Last edited:
Just an FYI, SimBY posted it, but the Starswarm stress test has been updated, i just done 4 runs on it for comparison and posted them in the Starswarm thread

http://forums.overclockers.co.uk/showthread.php?t=18576886&page=8

Interestingly OC'ing the core on the GPU made some difference, OC'ing the mem on the GPU had little impact at first glance, however i think the card processed more batches and stuff so infact was probably working harder for a negligable fps gain, im impressed.
 
You must log in or register to reply here.
Back
Top Bottom