• 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.

BF4 Retail CPU scaling measured

Fact is, my i5 2500k could only manage about ~60fps in intensive scenes/looking at a lot of the map in Shanghai. My 3770k does more like 75-80. The IPC and 0.2ghz bump over my 2500k would only account for a 10-15% increase but i'm getting more like a 25-30% increase so the hyperthreading is definitely doing something right. Even with the 3770k my GPU's still aren't being used at 100% on a full 64 player multiplayer map.

I'm not saying hyperthreading isn't helping, it surely is doing a lot of good.

I am saying an I5 Haswell with a mild OC will not be bottlenecked by any single GPU.
 
Just the way it is with Battlefield. I don't think I have seen anyone who gets 100% on both cards in an sli/xfire configuration on a full 64 player map in BF4 ( or 3 for that matter).

Single player both GPU's run at 95%+ utilisation.

Well my brother gets 100% each card, my friend gets 100% with 2x680's in high settings and I get 100% each card in ultra Jono.

Are you running 680's in sli?
 
Well my brother gets 100% each card, my friend gets 100% with 2x680's in high settings and I get 100% each card in ultra Jono.

Are you running 680's in sli?

670's. I find it hard to believe you are getting a constant 100% usage on both gpu's in a multiplayer match?
 
670's. I find it hard to believe you are getting a constant 100% usage on both gpu's in a multiplayer match?

Well believe it dude.

Flat out 100% on both gpu's with vsyn off and 47-49% with vsyn on.

It might be an nvidia driver thing. But however you look at it you've got an extra 10% somewhere in there to be found. (5% per card)

How is the 2Gb vram and 256bit memory bus holding up on the 670's?

Can you play on ultra with 4xmsaa as my BF4 m8 uses high 2x680's) to avoid hitting the vram frame buffer. I hit 2.5Gb's Vram sometimes how much are you using?
 
Well believe it dude.

Flat out 100% on both gpu's with vsyn off and 47-49% with vsyn on.

It might be an nvidia driver thing. But however you look at it you've got an extra 10% somewhere in there to be found. (5% per card)

How is the 2Gb vram and 256bit memory bus holding up on the 670's?

Can you play on ultra with 4xmsaa as my BF4 m8 uses high 2x680's) to avoid hitting the vram frame buffer. I hit 2.5Gb's Vram sometimes how much are you using?

I play at 1080p Ultra preset (so 4xMSAA) and performance is brilliant. No vram related issues at all. Not sure why your mate has to run on high? Also that might be why he is getting 100% usage. It is the ultra mesh that is a killer on the cpu apparently.

I still don't believe you are getting 100% usage on two cards in a full 64 player server though :p. What are your specs?
 
I play at 1080p Ultra preset (so 4xMSAA) and performance is brilliant. No vram related issues at all. Not sure why your mate has to run on high? Also that might be why he is getting 100% usage. It is the ultra mesh that is a killer on the cpu apparently.

I still don't believe you are getting 100% usage on two cards in a full 64 player server though :p. What are your specs?

Believe it I'm not talking ball-ox its how it is. Gain nothing by bullsitting.


Hes got a benq 120Hz monitor so he gets around 120fps or so when in high.

My specs are 3570k 4.6Ghz 2200Mhz cas10 memory, 2x7950's 1200Mhz and 1500Mhz memory. My AMD drivers are 13.8.

I'm sure others will confirm they get 100% on both gpu's also.

64, 48 or even 32 man servers use 100% irrespective. Remember that with vsyn off both gpu's try to push as many frames as is possible, therefore this is independent of no of people playing. BF4 has a 200fps limit which i dont get near.

64 man servers really hit the cpu a lot harder than 48 or 32, but both gpu's are fully utilised.

I'll triple check the gpu's again tonight just to be sure I'm not reading fan speed or someother gpu related factoid, but I'm sure its 100% on both cards...Still your making me doubt my sanity now Jono.
 
Last edited:
I still wont believe you unless I see it running with my own eyes with both gpus getting 100% :P

For example

http://www.youtube.com/watch?v=f80FpnEQFn8

Battlefield 3 with 780's in sli and a 4.6ghz 3770k. Gpu usage is nowhere near at a constant 100%. More like 60% in the outdoors bits.

Setter a member on here was running his 4770k at stock and was only getting 55% gpu usage on both his 670's. He posted the afterburner graphs no here.
 
I still wont believe you unless I see it running with my own eyes with both gpus getting 100% :P

For example

http://www.youtube.com/watch?v=f80FpnEQFn8

Battlefield 3 with 780's in sli and a 4.6ghz 3770k. Gpu usage is nowhere near at a constant 100%. More like 60% in the outdoors bits.

Well unless there is a driver issue or cpu bottleneck and/or vsyn is on then gpu usage should'nt be restricted to 60%.

Another scenario is if the cards were extremely powerful and hit 200fps and with say 60% gpu usage then yes its possible but 60% is an artificial limit or restiction.

In my mind gpu's with vsyn off always tend to max out unless vsyn is applied.

If the cpu limits perfomance than regardless gpu usage will be lower and fps will suffer. (just like you experienced).

If I get time tonight after work and father duties I'll take a screen grab..

As posted above, I'll double check for my sanity to confirm it again also....(Pray I havn't got dementia) :)
 
I still wont believe you unless I see it running with my own eyes with both gpus getting 100% :P

For example

http://www.youtube.com/watch?v=f80FpnEQFn8

Battlefield 3 with 780's in sli and a 4.6ghz 3770k. Gpu usage is nowhere near at a constant 100%. More like 60% in the outdoors bits.

Setter a member on here was running his 4770k at stock and was only getting 55% gpu usage on both his 670's. He posted the afterburner graphs no here.

With Vsyn on.
 
Well unless there is a driver issue or cpu bottleneck and/or vsyn is on then gpu usage should'nt be restricted to 60%.

Another scenario is if the cards were extremely powerful and hit 200fps and with say 60% gpu usage then yes its possible but 60% is an artificial limit or restiction.

In my mind gpu's with vsyn off always tend to max out unless vsyn is applied.

If the cpu limits perfomance than regardless gpu usage will be lower and fps will suffer. (just like you experienced).

If I get time tonight after work and father duties I'll take a screen grab..

As posted above, I'll double check for my sanity to confirm it again also....(Pray I havn't got dementia) :)

No vysnc, his fps is up in the top left corner.
 
Overclock your 8320 more. At 4.6GHz you are barely breaking a sweat. 4.8GHz is the sweetspot and you have the headroom in terms of voltage so why not use it. I would also run some formal benchmarks to get objective figures around the "doesn't run quite as well" statement.

Heard this before in BF4 and I don't find this to be the case on my 8320 paired with a 7990 (which eats a single 780 fairly comprehensively).

Edit: my bench figures are earlier in this thread for comparison.

I'll give it a try.
 
Guys and gals

The only way to get 99% always in BF4 is to

1. Ensure renderaheadlimit is set to 3 or above
2. You use RadeonPro or other NV software to force AA to max, Ambient Occlusion to Ultra, Mode to Supersampling etc etc. I did some testing and found my usage went up
3. Do not force it to 60fps or anything, use something higher like 120/144 - helps to have a monitor suitable for this

If I have it on High/Ultra in game, my usage is between 50 and 80%. With the above, hits 99% always.

I'll post a video
 
Guys and gals

The only way to get 99% always in BF4 is to

1. Ensure renderaheadlimit is set to 3 or above
2. You use RadeonPro or other NV software to force AA to max, Ambient Occlusion to Ultra, Mode to Supersampling etc etc. I did some testing and found my usage went up
3. Do not force it to 60fps or anything, use something higher like 120/144 - helps to have a monitor suitable for this

If I have it on High/Ultra in game, my usage is between 50 and 80%. With the above, hits 99% always.

I'll post a video

Thats 99% with CF280x's?

To be honest with you I never really thought much about this before. I thought 100% gpu usage on both cards was typical but clearly some people dont get high scaling from 1 to 2 cards. I always use custom ultra with 4xmsaa with all settings maxed out @ 1920x1200.

I have disabled frame pacing and UDLP and thats it, Just plug the card in and attached cross brigde, I didn't even install new drivers and there it was. (Plus a 1000W psu )

I can't wait to get a third 7950 gpu to see my i5 scream in cpu bottleneck pain.

How are the new beta drivers? what fps are you getting?
 
Last edited:
Thats 99% with CF280x's?

To be honest with you I never really thought much about this before. I thought 100% gpu usage on both cards was typical but clearly some people dont get high scaling from 1 to 2 cards. I always use custom ultra with 4xmsaa with all settings maxed out @ 1920x1200.

I have disabled frame pacing and UDLP and thats it, Just plug the card in and attached cross brigde, I didn't even install new drivers and there it was. (Plus a 1000W psu )

I can't wait to get a third 7950 gpu to see my i5 scream in cpu bottleneck pain.

How are the new beta drivers? what fps are you getting?

Yes, 2 x 280Xs

I feel jelous of you mate. I have always had to mess and prod and play with my GPUs to achieve maximum usage in the BF series.
 
Thats 99% with CF280x's?

To be honest with you I never really thought much about this before. I thought 100% gpu usage on both cards was typical but clearly some people dont get high scaling from 1 to 2 cards. I always use custom ultra with 4xmsaa with all settings maxed out @ 1920x1200.

I have disabled frame pacing and UDLP and thats it, Just plug the card in and attached cross brigde, I didn't even install new drivers and there it was. (Plus a 1000W psu )

I can't wait to get a third 7950 gpu to see my i5 scream in cpu bottleneck pain.

How are the new beta drivers? what fps are you getting?

But that is simply because he is increasing the settings until the framerate become low enough for the GPU to be the bottleneck again.

I could get my GPU usage to maximum 100% as well if I went with loads of supersampling or put the resolution scale above 100% in the BF4 options. Thing is I prefer the 100fps over 60fps :p
 
OC'd the 8320 to 4.8, even tried 5GHz. But it just can't maintain 120 FPS. It will stick around 100 on some lower settings but the 3570k can maintain 120 flawlessly even with AA @ 1080p.

I don't have any formal benchmarks or anything, just going by how the game feels and FPS readouts.
 
But that is simply because he is increasing the settings until the framerate become low enough for the GPU to be the bottleneck again.

I could get my GPU usage to maximum 100% as well if I went with loads of supersampling or put the resolution scale above 100% in the BF4 options. Thing is I prefer the 100fps over 60fps :p

I'm not increasing the resolution scaling or using super sampling its set to 100%

I just used ultra 4xmsaa, 1920x1200, high occulsions. fps 120-140 with dips to 90fps.

The lowest i've ever seen is 70fps in Paracel Storm.

Yes, 2 x 280Xs

I feel jelous of you mate. I have always had to mess and prod and play with my GPUs to achieve maximum usage in the BF series.

Its probably more to do with AMD sorting out the crossfire drivers in the previous 6 months.

How high have you oc'd those cards?
 
Setter a member on here was running his 4770k at stock and was only getting 55% gpu usage on both his 670's. He posted the afterburner graphs no here.
Tbh, ab isnt ideal for checking usage. My average card usage can actually get lower. The pic below shows both cpu/and gpu usage after a full round of 64 player cq.

dY1Dnk6.png

Max cpu usage does get pretty high, but then i am running at stock. Cards max at 96% but as you can see the average is much lower. Ill be reclocking the cpu soon as i got my replacement ram in. Will leave hwinfo running and check results again.
 
Last edited:
Ive just swapped an i5 2500k (4.5ghz) for an i7 3770K (4.5Ghz) and I'm still not seeing very high GPU usage with my sli 670's. More like 60/70%.
 
Ive just swapped an i5 2500k (4.5ghz) for an i7 3770K (4.5Ghz) and I'm still not seeing very high GPU usage with my sli 670's. More like 60/70%.

Really? I have seen quite a jump. GPU usage still isn't maxed out but it averages about 85% now rather than about 65%
 
Back
Top Bottom