***THE BF4 BENCHMARKS THREAD***

[gpuerrilla]

Getting 50+fps with stock CPU and mantle on 14.2 - not bad!

 

[Rroff]

The fps figures in the second pic look way too low for a 780ti, im getting much better on a bog standard 780 using the same drivers. Im guessing the runs were done with resolution scale turned to 150% or thereabouts.

I questioned it before - with those settings, on a full 64 CQ large server even my min fps are higher than their average - even removing my CPU overclock and limiting my 780 to reference clocks - infact even video recording at the same time only just about drops my min to their average.

I find both sets of benchmarks hes posted seem well off for a 780 or 780ti actual results.

 

[humbug]

I questioned it before - with those settings, on a full 64 CQ large server even my min fps are higher than their average - even removing my CPU overclock and limiting my 780 to reference clocks - infact even video recording at the same time only just about drops my min to their average.

I find both sets of benchmarks hes posted seem well off for a 780 or 780ti actual results.

What is it that you want me to say, that they are lying?

Do you think they are lying?

 

[setter]

What is it that you want me to say, that they are lying?

Do you think they are lying?

Yes they are lying mate, my stock wf 780 averages 80 fps at 1920x1200, Rroff runs a ghz wf 780 on a 100mhz faster stock boost than my card so his fps are a wee bit higher. From testing bf4 with 150% res scale i know that the figures in that pic are using that setting to skew the results. Bf4 benched at 1920x1200, 150% res scale, all ultra but x2 msaa.

Frames, Time (ms), Min, Max, Avg
37612, 600000, 35, 82, 62.687

Those benchmarks are about as useful as a chocolate fireguard/kettle/ashtray on a motorbike.

 

[Rroff]

What is it that you want me to say, that they are lying?

Do you think they are lying?

Dunno about lying but something isn't right with their results especially the siege of shanghai with 1920x1080 ultra one. (Unless my 4820 and heavily tweaked RAM is giving that much of an advantage which I doubt).

 

[setter]

I can average 80fps on a stock gtx 780, (well slight factory oc from gigabyte) on shanghai, ultra, x4 msaa, the other aa setting on high, 16xaf. HBAO on and blur which i hate, (migraine issue). In those results, i wonder who sponsored/payed for them because theyre laughable.

4770k and a 4820k are pretty similair even when oc'd, which wont make an awful difference in bf4.

 

[humbug]

Yes they are lying mate, my stock wf 780 averages 80 fps at 1920x1200, Rroff runs a ghz wf 780 on a 100mhz faster stock boost than my card so his fps are a wee bit higher. From testing bf4 with 150% res scale i know that the figures in that pic are using that setting to skew the results. Bf4 benched at 1920x1200, 150% res scale, all ultra but x2 msaa.

Frames, Time (ms), Min, Max, Avg
37612, 600000, 35, 82, 62.687

Those benchmarks are about as useful as a chocolate fireguard/kettle/ashtray on a motorbike.

I think you and Roff are being a bit silly.

Why would they lie? i average 70 FPS on DX @ 1080P on Ultra minus MSAA just randomly playing the game, thats a lot higher than what they have for the 290X.

Yet i also know there are parts on all maps where my FPS can drop to 40 FPS, this on a 4Ghz Bloomfield i7 and the x6. as i have illustrated here http://forums.overclockers.co.uk/showpost.php?p=25967192&postcount=6964 Others with newer i7's all experience the same thing. That which you keep ignoring, the amount of times me and others have displayed this in here and yet seemingly it never enters into your calculations when you write silly things like that, its as if your blind to it.
If i run up and down that road on Shanghai all i get is ~50 FPS at best, same between flag A and B on Paracel Storm, i could go on....

So me and them also must also be liars, and frankly i take offence to that.

What they did here and the results are absolutely in line with what people who can use Mantle experience.

Your ignoring the fact that CPU bottlenecks exist in BF4 DX and are then making a blanket accusation about something you know absolutely nothing about.

 

[Rroff]

^^ My point is I can use the settings they supposedly used, on the same maps and try and find the worst possible scenario and I'm still not even close to as low fps as they are getting so whatever is going on something ain't right and I'm not alone in that several people have backed up similiar numbers from their 780s let alone 780tis.

I have to be video capturing in the most active parts of the map to get even close to as low numbers as they have.

For instance if I try to find the most DX inefficent part of siege of shanghai with the settings and res they use on a full 64 player server with a lot of players in that area I don't drop below ~67fps which is still above their average let alone min and my average fps running around the map playing is almost double their average and thats with a 780 albeit a heavily clocked one let alone a 780ti.

 

[setter]

Humbug, i wouldnt say i or Rroff are being silly or offensive to you and if it came across that way i apologise. But the fps figures in that graph are way way too low for a gtx 780ti, let alone a gtx 780. The fps figures i posted above were from a run at 1920x1200, ultra and x2 msaa but with 150% res scale. My fps figures doing so are very close to those in that graph. At 100% res scale and x4msaa on ultra with HBAO on, x16af im averaging 80fps on a stock gtx 780 that boosts to a rather poor 1097mhz. Few benches ive run lately, bf4 on the highest possible settings at 1920x1200.

Frames, Time (ms), Min, Max, Avg
15791, 154987, 62, 171, 101.886

Frames, Time (ms), Min, Max, Avg
50382, 600000, 46, 144, 83.970

Frames, Time (ms), Min, Max, Avg
33442, 396664, 47, 164, 84.308

Frames, Time (ms), Min, Max, Avg
49815, 600000, 53, 147, 83.025

Theese are all from shanghai and paracel 64 player cq large servers, 60 players minimum when i joined. Max settings at my rather ancient res of 1920x1200. I only bench theese two maps as theyre the most demanding.

 

[humbug]

^^ My point is I can use the settings they supposedly used, on the same maps and try and find the worst possible scenario and I'm still not even close to as low fps as they are getting so whatever is going on something ain't right and I'm not alone in that several people have backed up similiar numbers from their 780s let alone 780tis.

I have to be video capturing in the most active parts of the map to get even close to as low numbers as they have.

For instance if I try to find the most DX inefficent part of siege of shanghai with the settings and res they use on a full 64 player server with a lot of players in that area I don't drop below ~67fps which is still above their average let alone min and my average fps running around the map playing is almost double their average and thats with a 780 albeit a heavily clocked one let alone a 780ti.

Humbug, i wouldnt say i or Rroff are being silly or offensive to you and if it came across that way i apologise. But the fps figures in that graph are way way too low for a gtx 780ti, let alone a gtx 780. The fps figures i posted above were from a run at 1920x1200, ultra and x2 msaa but with 150% res scale. My fps figures doing so are very close to those in that graph. At 100% res scale and x4msaa on ultra with HBAO on, x16af im averaging 80fps on a stock gtx 780 that boosts to a rather poor 1097mhz.

After pages and pages of explanations your still not getting it.

"oh its lower than i get on my 780/TI" where you test it i'm sure it is, not where they test it, and not where i tested it. CPU bottlenecks have absolutely nothing to do with what GPU your running, forget the GPU, if you test it in CPU bottleneck parts of the game you might as well be running a 7850 or a GTX 660, you will get the same results that you do on a 780TI

You keep taking about this and that GPU. how much its overclocked.... its irrelevant, you could have it running at 1500Mhz and you will get the same result.

The test is how much does the CPU hold the GPU back, in DX its twice as much as in Mantle, thats why their and my Mantle results are twice that of DX.

This is really basic stuff guys.

 

[Rroff]

I'm literally saying that I can hunt out the worst, CPU bound scenario on that map, I can stay there entirely and benchmark just that bit and I'm still well above their results which is why I question it - using the same supposed settings and so on even with video capturing I'm ~36% higher average fps than they get let alone if I run around a larger scope of the map. And anyone with a properly working 780 based setup on a decent i7 should see the same thing let alone a 780ti.

 

[setter]

After pages and pages of explanations your still not getting it.

"oh its lower than i get on my 780/TI" where you test it i'm sure it is, not where they test it, and not where i tested it. CPU bottlenecks have absolutely nothing to do with what GPU your running, forget the GPU, if you test it in CPU bottleneck parts of the game you might as well be running a 7850 or a GTX 660, you will get the same results that you do on a 780TI

You keep taking about this and that GPU. how much its overclocked.... its irrelevant, you could have it running at 1500Mhz and you will get the same result.

The test is how much does the CPU hold the GPU back, in DX its twice as much as in Mantle, thats why their and my Mantle results are twice that of DX.

This is really basic stuff guys.

Iirc you have a bloomfield i7 1136 cpu. Ive owned several of those, both clocked pretty highly. 4.2ghz on my old 920, 4.3 on the 930 that i keep as a download box. With the 780 at stock from my rig in sig in the 930 rig, bf4 plays the same. The only way the 930 fell behind was when running multi gpu, (sli 670's) despite it being a dual x16 pcie board compared to dual x8 pcie, the same cards were ran in a 3770k at 4.5ghz, (sold as it wasnt a good clocker) and in my current 4770k setup, which clock for clock is about 300mhz faster than the old ivy chip.

 

[humbug]

I'm literally saying that I can hunt out the worst, CPU bound scenario on that map, I can stay there entirely and benchmark just that bit and I'm still well above their results which is why I question it - using the same supposed settings and so on even with video capturing I'm ~36% higher average fps than they get let alone if I run around a larger scope of the map. And anyone with a properly working 780 based setup on a decent i7 should see the same thing let alone a 780ti.

Replicate this with your i7 4820K. put the sceenshot up here.

Its very simple stuff. it will only take a few minutes.

I'm off to bed. looking forward to your response tomorrow.

 

[Rroff]

Hang around I've got one I did from there recently.

EDIT: Actually will get a new one as theres been some performance enchancements in both game and drivers since I took the last screenshot from that position.

This is one of the ones I did previously ( http://i.imgur.com/rh0SRC7.jpg ) - its upto around 75fps with latest drivers/patch IIRC. (Full Ultra settings at 1920x1080).

 

[setter]

Replicate this with your i7 4820K. put the sceenshot up here.

Its very simple stuff. it will only take a few minutes.

I'm off to bed. looking forward to your response tomorrow.

What settings? And mp or sp?

 

[tommybhoy]

Fair enough. Tommybhoy made out everything was sorted but thanks for the honest answer

Greg, can't believe your now putting up blatant unfounded lies, no need to start putting up lies while at the same time attempting to make me look the dishonest one.

Where did I mention BF4 Mantle being sorted?

The majority enjoying Mantle is what I stated, all I did was challenge one of your many 'Mantle's broken for most' statements.

Now I'm a liar, get a grip, seriously.

Well BF4 still isn't working correctly for most with Mantle.

Why do you keep banging the drum about a broken Mantle when the majority are enjoying it?

The only poll I know of regarding Mantle:

Lets break it down, 2723 AMD users:

1331 loved it
847 haven't tried it and are waiting on a more refined version
423 went back to DX
121 couldn't install the driver properly

More than a 3:1 ratio of Mantle working for most.

http://www.techpowerup.com/

Refrain from the fud bud.

http://forums.overclockers.co.uk/showpost.php?p=25964693&postcount=180

Not sure how many times I have to say it. I hope Mantle is a success and if it is working well for all, fantastic.

Doesn't matter how many times you say you hope Mantle is a success in one hand when you negate it by stating it doesn't work for most with every other breath, it's fud reporting while not even using mantle:

Well BF4 still isn't working correctly for most with Mantle.

I'm under no illusions some haven't had a far from positive Mantle experience, but it certainly aint the majority with the numbers provided.

Now explain why AMD users are turning settings down, as I thought the idea was to have higher frame rates so you could add to the details?

Obvious answer is it's the minority having problems going by the consensus.

I've never had to turn anything down, I can turn settings up(res scaling) and play on lesser fps than DX and it's still a better experience than DX, it's like a lavish smothering of salty Lurpak-the full flavour you can only taste first hand.

If it is working correctly, then I apologise and good stuff.

You don't need to apologise, just take notice of the positive experiences for a change instead of only absorbing the negatives-which ends up equating to your jelly belly misinforming the masses regarding Mantle.

http://forums.overclockers.co.uk/showpost.php?p=25964819&postcount=185

I simply pointed out that Mantle works for the majority, yet you keep informing that it's broken for the majority, nothing more nothing less.

There is only one person being dishonest about this situation and it aint me, and it's clear reading for everyone to see where the dishonesty stems from here.

Please refrain from attempting to make me look dishonest/putting doubt on my credibility in future please, before this, I have never accused you of dishonesty before-or anyone else for that matter.

 

[setter]

Well, heres a few screens from shanghai, 48 players on a cq 64 man server. My settings first.

Spoiler

Ingame shot.

Spoiler

Swedish server and quite laggy.

 

[chief barker]

@ Tommy

Someone needs to justify their £1600 worth of GPUs

 

[whyscotty]

4770K @ stock Titan @ stock 64 player server

Spoiler

 

[Silent_Scone]

That chap has 35" of d*ck apparently. Always something going on in chat