***THE BF4 BENCHMARKS THREAD***

[Gregster]

Calm down mate no one is ripping you to pieces, I thought we were just having a debate about frame variance. No one is trying to say you're video is not smooth and your average fps look great. I was only trying to explain why the pcper FCAT results show spikes as i see the same thing as your video shows on DX. That's not to say you're having a bad experience at all, which I've tried to make clear.

I am perfectly calm but posting a vid and then getting told it isn't smooth because a green line isn't level, even though nothing in the vid shows anything different and then getting told to bench a 64 man CQ server and do this and that, just to please you is going a bit overboard don't you think? You called me Greggles which is condescending and rude.

I am all up for a reasoned debate about anything but to be told this and that but yet no one can tell me where it looks like it is stuttering and hitching and only a pic of my game with a small spike (which incidentally doesn't show at all in the smoothness when I replay it at that part) and all I did was post a vid of my frames and smoothness using ShadowPlay, I am not interested.

 

[Rusty0611]

It's all well and good comparing a green line but if the minor differences in the line result in no discernible difference in end user experience then so what right?

 

[shankly1985]

I am perfectly calm but posting a vid and then getting told it isn't smooth because a green line isn't level, even though nothing in the vid shows anything different and then getting told to bench a 64 man CQ server and do this and that, just to please you is going a bit overboard don't you think? You called me Greggles which is condescending and rude.

I am all up for a reasoned debate about anything but to be told this and that but yet no one can tell me where it looks like it is stuttering and hitching and only a pic of my game with a small spike (which incidentally doesn't show at all in the smoothness when I replay it at that part) and all I did was post a vid of my frames and smoothness using ShadowPlay, I am not interested.

LOL
Really?
We started off by saying how much smoother Mantle it vs DX. I even agreed with you DX is smooth, you no any different till you tried out Mantle.

I then give you a reason, by showing you have spikes in real time. Great if you dont notice them, dont try because if you did then that would be a real shame.
All I was saying is Mantle is much better the DX frame rate and frame timing.

The point about showing TDM is a given fact you not showing us what a true BF4 test should be like using the game to its fullest.
TDM dont stress a system like 64 player, you then showed us a video standing far out side the map. Again not stressing anything, thats just like looking up at the sky and watching your frame rate hit 200fps

 

[LtMatt]

I am perfectly calm but posting a vid and then getting told it isn't smooth because a green line isn't level, even though nothing in the vid shows anything different and then getting told to bench a 64 man CQ server and do this and that, just to please you is going a bit overboard don't you think? You called me Greggles which is condescending and rude.

I am all up for a reasoned debate about anything but to be told this and that but yet no one can tell me where it looks like it is stuttering and hitching and only a pic of my game with a small spike (which incidentally doesn't show at all in the smoothness when I replay it at that part) and all I did was post a vid of my frames and smoothness using ShadowPlay, I am not interested.

No one has said your video is not smooth though. I only commented that i get similar gpu spikes under DX too and that these likely indicate hitching and stutter on the pcper FCAT results, something which is not present under Mantle. That's just my opinion of it and our DX results look similar to me. I'm happy with my DX results and so should you be, that's all you've tried.

All I tried to say about the youtube video is tiny hitching or stuttering (if present) is not going to show up in a 30fps video. That's why people use 60fps or higher videos for slow motion.

If you can't detect any stutter, that's great. I couldn't either before i tried Mantle and i was blissfully ignorant to it as well. Sometimes that's the best way to be i think. As I've already said I've just benched DX this morning and was quite happy with the performance and smoothness on offer. No one is saying your video is not smooth and neither is your gameplay experience. All we're saying is, it can get so much better.

Sorry for calling you Greggles. Saw someone else call you it and thought it was a nickname. I got told off for calling Rroff Rroffles once too. I like to give people harmless nicknames. Hell even Rusty is now known as Rusters.

 

[pgi947]

Again there seems to be some confusion.

A slight spike in the green GPU line doesn't mean its stuttering....at all, it means the rendered frame is being delivered slower then the previous, which depending on refresh and frame rate it occoured at may be completely unnoticeable. You need to look at the relative graph scale too, a big spike might only represent 3ms uplift in turn meaning a drop 10fps or so? That won't cause stutter at the frame rates Greg is at. Mantle remains flat as each frame is at a consistent rate where as DX on both sides of the fence have more variable frame rates, in turn a less then perfectly flat line.

Frame times are used to find runt frames, a frame being rendered at the correct speed but being delivered late, a 16.6ms frame being delivered at 33ms for example. THIS is what causes stutter and what kicked off the whole frame time argument. You will get frame reduction stutter, like dropping from 60 to 20, there would be a spike in the graph but it wouldn't represent a frame time causing the issue, it would merely be the reduction in rate

 

[LtMatt]

Again there seems to be some confusion.

A slight spike in the green GPU line doesn't mean its stuttering....at all, it means the rendered frame is being delivered slower then the previous, which depending on refresh and frame rate it occoured at may be completely unnoticeable. You need to look at the relative graph scale too, a big spike might only represent 3ms uplift in turn meaning a drop 10fps or so? That won't cause stutter at the frame rates Greg is at. Mantle remains flat as each frame is at a consistent rate where as DX on both sides of the fence have more variable frame rates, in turn a less then perfectly flat line.

Frame times are used to find runt frames, a frame being rendered at the correct speed but being delivered late, a 16.6ms frame being delivered at 33ms for example.

The frame rate can change drastically on Mantle and it will not spike. The same cannot be said for DX. Although not every spike will mean stutter, the instances of stutter will be occurring when the spikes do. With frames being rendered at different speeds, this is the main cause of microstutter, hitching whatever you want to call it. It's then up to the gpu to manage that frame time delivery using pacing. However as we can see by looking at AMD/Nvidia DX FCAT results this is impossible to manage without any stutter. High framerates surely help mind and definitely minimise the perceived stutter, but it is most definitely still present. For AMD the higher the resolution goes, the better their frame consistency gets using FCAT in BF4, Nvidia seem to be the opposite. Maybe a difference in how they approach this idk.

 

[shankly1985]

The frame rate can change drastically on Mantle and it will not spike. The same cannot be said for DX. Although not every spike will mean stutter, the instances of stutter will be occurring when the spikes do. With frames being rendered at different speeds, this is the main cause of microstutter, hitching whatever you want to call it. It's then up to the gpu to manage that frame time delivery using pacing. However as we can see by looking at AMD/Nvidia DX FCAT results this is impossible to manage without any stutter. High framerates surely help mind and definitely minimise the perceived stutter, but it is most definitely still present.

Yeah! you can see from my video I get massive frame drops but still the Render line runs flat. DX this would be completely different. The bigger the frame change the more spikes I would get.

 

[humbug]

The trouble with this statement is variety. Mantle is in BF4 and Thief and as far as Thief goes, when you get to 1440P resolution and higher, Mantle starts to lose out or run a single card and again, Mantle starts to lose out, even at 1080P. You only have to look at the Thief thread you run for evidence of that or ask Kaap/read Kaaps findings. He has had problems in Mantle games with VRAM at 4K. The same could be said for BF4 but as we only have reviewers to rely on, everything can be taken with a pinch of salt. Check out several different reviewers for several different results and there is a PCPer article that gets bandied about quite a bit showing nVidia in a bad light. Sure I don't have a problem with that but if you look a page or so back, my video of a 30+ minute game does show how smooth the game runs.

I have said it several times about being happy if nVidia adopted Mantle and I seriously mean it. I feel less CPU usage and higher GPU usage would be great. I doubt it would happen but nVidia do have gamers interests to consider and it wouldn't do them any harm for PR reasons if they took it on.... That is of course if Mantle does become a truly open source API. I just wish there was a bench for BF4 to see who is fastest. I honestly don't care but it would nip a lot of arguments early on

Spoken like someone who does not have Mantle. look at this.

#1, LtMatt: 290P @ 1250/1625, 2700K @ 5.3Ghz - Mantle FPS: Min 109.5 / Avr 142.5
#2, Gregster: Titan @ 1267/3758, 3930K @ 4.6Ghz - DirectX FPS: Min 72.8 / Avr 130.2

Crossfire and SLI With a much weaker CPU Matt has 40% higher Min FPS than you, because he's using Mantle.

@ 1440P he is still 10% ahead of you with that same weaker CPU.

#1, Gregster: Titan @ 1267/3758, 3930K @ 4.6Ghz - DirectX FPS: Min 76.9 / Avr 107.9 Link
#2, LtMatt: 290P @ 1250/1625, 2700K @ 5.3Ghz - Mantle FPS: Min 84.2 / Avr 102.4

i also watched your BF4 video Greg, i did not want to say anything at the time as you so proudly proclaimed its smoothness, it is not, its a bit of a mess to be frank, there is significant stuttering recorded in the line draw graph and its very visible to my eye's in footage. its not good, its a stuttering mess.

Mantle isn't just about FPS, its about consistency in FPS, its about not dropping Frames, its about stringing those frames together much better than DX does, Mantle is much smoother, after using it for 6 moths your BF4 DX Video makes me cringe as its jerking and stuttering around, i couldn't even play it like that anymore.

No one has said your video is not smooth though.

Oh i just did, lets not beat around the bush, he says Mantle has no real benefit, with that footage? he has no idea.

 

[Gregster]

LOL
Really?
We started off by saying how much smoother Mantle it vs DX. I even agreed with you DX is smooth, you no any different till you tried out Mantle.

I then give you a reason, by showing you have spikes in real time. Great if you dont notice them, dont try because if you did then that would be a real shame.
All I was saying is Mantle is much better the DX frame rate and frame timing.

The point about showing TDM is a given fact you not showing us what a true BF4 test should be like using the game to its fullest.
TDM dont stress a system like 64 player, you then showed us a video standing far out side the map. Again not stressing anything, thats just like looking up at the sky and watching your frame rate hit 200fps

No, I posted a vid and you responded by telling me it wasn't the best map/format to showcase and then told me to get on a more demanding map with 64 players. I wasn't posting for the benefit of it getting ripped to bits, I posted it showing how ShadowPlay works so well and frames/smoothness were good. I have uploaded a few vids recently having just got Fibre and quite pleased with how they came out. I detest CQ games, as anyone who games with me will tell you and much prefer TDM/Domination and what is the point of me posting anything else either, I will just get told I don't know any different.

I am a proponent of Mantle and sing the praises of it but to be told that I need to run this or that and do this and do that is petty and unwarranted. I have DX, I dare say Mantle is smoother but do I need it ramming down my throat at every opportunity? I don't think so.

Edit:

Cheers Humbug. I guess I will have to put up with the stuttering mess that I am getting and please tell me where I have said that Mantle has no real benefit?

 

[Geeman1979]

Spoken like someone who does not have Mantle. look at this.

#1, LtMatt: 290P @ 1250/1625, 2700K @ 5.3Ghz - Mantle FPS: Min 109.5 / Avr 142.5
#2, Gregster: Titan @ 1267/3758, 3930K @ 4.6Ghz - DirectX FPS: Min 72.8 / Avr 130.2

Crossfire and SLI With a much weaker CPU Matt has 40% higher Min FPS than you, because he's using Mantle.

@ 1440P he is still 10% ahead of you with that same weaker CPU.

#1, Gregster: Titan @ 1267/3758, 3930K @ 4.6Ghz - DirectX FPS: Min 76.9 / Avr 107.9 Link
#2, LtMatt: 290P @ 1250/1625, 2700K @ 5.3Ghz - Mantle FPS: Min 84.2 / Avr 102.4

i also watched your BF4 video Greg, i did not want to say anything at the time as you so proudly proclaimed its smoothness, it is not, its a bit of a mess to be frank, there is significant stuttering recorded in the line draw graph and its very visible to my eye's in footage. its not good, its a stuttering mess.

Mantle isn't just about FPS, its about consistency in FPS, its about not dropping Frames, its about stringing those frames together much better than DX does, Mantle is much smoother, after using it for 6 moths your BF4 DX Video makes me cringe as its jerking and stuttering around, i couldn't even play it like that anymore.



Oh i just did, lets not beat around the bush, he says Mantle has no real benefit, with that footage? he has no idea.

He's running virtually the same setup as I do and I guarantee you he's not seeing stutter.

 

[Ian Evey]

No, I posted a vid and you responded by telling me it wasn't the best map/format to showcase and then told me to get on a more demanding map with 64 players. I wasn't posting for the benefit of it getting ripped to bits, I posted it showing how ShadowPlay works so well and frames/smoothness were good. I have uploaded a few vids recently having just got Fibre and quite pleased with how they came out. I detest CQ games, as anyone who games with me will tell you and much prefer TDM/Domination and what is the point of me posting anything else either, I will just get told I don't know any different.

I am a proponent of Mantle and sing the praises of it but to be told that I need to run this or that and do this and do that is petty and unwarranted. I have DX, I dare say Mantle is smoother but do I need it ramming down my throat at every opportunity? I don't think so.

Edit:

Cheers Humbug. I guess I will have to put up with the stuttering mess that I am getting and please tell me where I have said that Mantle has no real benefit?

Maybe all of us nvidia users need glasses or something, I like the fact that no one can see a stuttery mess except humbug.

I would give up Gregster as even if you say bf4 is smooth it's not, and amd are the best regarding gaming, The sooner everyone buys an amd card the better.

 

[GoogalyMoogaly]

Spoken like someone who does not have Mantle. look at this.

#1, LtMatt: 290P @ 1250/1625, 2700K @ 5.3Ghz - Mantle FPS: Min 109.5 / Avr 142.5
#2, Gregster: Titan @ 1267/3758, 3930K @ 4.6Ghz - DirectX FPS: Min 72.8 / Avr 130.2

Crossfire and SLI With a much weaker CPU Matt has 40% higher Min FPS than you, because he's using Mantle.

@ 1440P he is still 10% ahead of you with that same weaker CPU.

#1, Gregster: Titan @ 1267/3758, 3930K @ 4.6Ghz - DirectX FPS: Min 76.9 / Avr 107.9 Link
#2, LtMatt: 290P @ 1250/1625, 2700K @ 5.3Ghz - Mantle FPS: Min 84.2 / Avr 102.4

i also watched your BF4 video Greg, i did not want to say anything at the time as you so proudly proclaimed its smoothness, it is not, its a bit of a mess to be frank, there is significant stuttering recorded in the line draw graph and its very visible to my eye's in footage. its not good, its a stuttering mess.

Mantle isn't just about FPS, its about consistency in FPS, its about not dropping Frames, its about stringing those frames together much better than DX does, Mantle is much smoother, after using it for 6 moths your BF4 DX Video makes me cringe as its jerking and stuttering around, i couldn't even play it like that anymore.



Oh i just did, lets not beat around the bush, he says Mantle has no real benefit, with that footage? he has no idea.

Weaker CPU?
The one that's clocked 700MHz faster?
Unless Thief is using more than 8 threads I'm not sure how it's a weaker CPU.

I've not noticed any benefit with Mantle as far as smoothness goes (doesn't mean I'm not getting any), what I do notice though is running a 120Hz monitor compared to 60Hz. If I cap my FPS to 60, even if it's a consistent 60 it feels awful, gives eye strain, there's ghosting and it isn't enjoyable.
I managed fine at 60Hz before and I'm sure there are people using 60Hz now that think it's fine but going back you notice these things. I suspect if you're used to 1440p going back to 1080p would look horribly blocky.

Maybe this will be another 'discussion' that can be put to bed when DX 12 is out (hopefully giving the same benefit to everyone, where it's noticed or not).
I seem to recall there was lots of talk about AMD and runt frames and issues with Crossfire back before the frame-pacing fix. But there were also people back then that didn't notice it or weren't bothered by it. There were probably similar arguments back then but with the sides reversed. If in the future Nvidia has better frametimes we'll probably have the same argument all over again though.

Why do frame times always seem to get brought up in the BF4 thread and not in Hitman: Absolution or Sleeping Dogs benchmark threads?
Presumably without Mantle these games are a mess on everyone's systems? So do people still play them?

 

[uksoldierboy]

Screw the stutter! Come on guys it's all good, I think if someone was to say Mantle makes no difference to the smoothness of play in BF4 then we could all agree that the person has either never used Mantle, or is talking out of their arse.

It does make a difference, i guess how much of a difference is down to the individual.

for me personally, it's very noticeable, switching between Mantle and DX11, but at the same time, if i was to play using DX11 for a few rounds I'm sure that I'd soon get used to it, to the point that i wouldn't even notice anymore.

Greg looking at the vid you posted, yes, that graph could be smoother, its got a lot of spikes in it, if you were to use a Mantle i have no doubt you'd feel the difference.

At the same time you're happy with it, that's all that matters, if you find it perfectly playable, awesome.

DX11 on BF4 is awesome, i was playing it fine before mantle was in place and enjoyed it.
Mantle Awesome-er, it makes a great difference, it makes the smooth gameplay on you'd get with DX11, feel even smoother.

Bring on DX12 and future improvement for Mantle, it's going to be great.

 

[LtMatt]

Screw the stutter! Come on guys it's all good, I think if someone was to say Mantle makes no difference to the smoothness of play in BF4 then we could all agree that the person has either never used Mantle, or is talking out of their arse.

It does make a difference, i guess how much of a difference is down to the individual.

for me personally, it's very noticeable, switching between Mantle and DX11, but at the same time, if i was to play using DX11 for a few rounds I'm sure that I'd soon get used to it, to the point that i wouldn't even notice anymore.

Greg looking at the vid you posted, yes, that graph could be smoother, its got a lot of spikes in it, if you were to use a Mantle i have no doubt you'd feel the difference.

At the same time you're happy with it, that's all that matters, if you find it perfectly playable, awesome.

DX11 on BF4 is awesome, i was playing it fine before mantle was in place and enjoyed it.
Mantle Awesome-er, it makes a great difference, it makes the smooth gameplay on you'd get with DX11, feel even smoother.

Bring on DX12 and future improvement for Mantle, it's going to be great.

Can't fault any of that. +1'd.

 

[Gregster]

It's all cool for me and I guess I am happy with my stuttering mess gameplay

Still curious to know where I have ever said that Mantle has no benefits but that won't be shown, as I have never said it.

 

[shankly1985]

No, I posted a vid and you responded by telling me it wasn't the best map/format to showcase and then told me to get on a more demanding map with 64 players. I wasn't posting for the benefit of it getting ripped to bits, I posted it showing how ShadowPlay works so well and frames/smoothness were good. I have uploaded a few vids recently having just got Fibre and quite pleased with how they came out. I detest CQ games, as anyone who games with me will tell you and much prefer TDM/Domination and what is the point of me posting anything else either, I will just get told I don't know any different.

I am a proponent of Mantle and sing the praises of it but to be told that I need to run this or that and do this and do that is petty and unwarranted. I have DX, I dare say Mantle is smoother but do I need it ramming down my throat at every opportunity? I don't think so.

Edit:

Cheers Humbug. I guess I will have to put up with the stuttering mess that I am getting and please tell me where I have said that Mantle has no real benefit?

Since when was I telling you what to do? I was asking or more like suggesting or pointing out that TDM is easy on the system or giving advice.

I was not TELLING you to do anything.

 

[Gregster]

Since when was I telling you what to do? I was asking or more like suggesting or pointing out that TDM is easy on the system or giving advice.

I was not TELLING you to do anything.

Not the best map to be show casing tbh TDM also takes less performance.
Get on another more demanding map with 64 players

 

[scotsdragon]

Just jump on the game shoot each other and get it over with oh wait that's WED im gunning for you all *evil laugh* mwuahahahahahaha lol

 

[Ian Evey]

Just jump on the game shoot each other and get it over with oh wait that's WED im gunning for you all *evil laugh* mwuahahahahahaha lol

You have a 780 though, It will be a stuttery mess

 

[uksoldierboy]

Just jump on the game shoot each other and get it over with oh wait that's WED im gunning for you all *evil laugh* mwuahahahahahaha lol