• 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.

306.97 WHQL release

BF3 is better for me in sli with these, high 70% to mid 80% usage now and seem to be holding the fps better than any others

Just saw what you typed... Isn't that like pretty bad? 98/99 usage is much better right? :) It's what I had on my SLI 480 with my system @ 4Ghz.

If you have 70 - 80% doesn't that mean your system can't give your GTX's as much as they need?
 
Just saw what you typed... Isn't that like pretty bad? 98/99 usage is much better right? :) It's what I had on my SLI 480 with my system @ 4Ghz.

If you have 70 - 80% doesn't that mean your system can't give your GTX's as much as they need?

i5's do limit SLI to high 80's usage unfortunately.

The extra threads on i7's do release the CPU bottleneck in MP64 in BF3.

I had an i5 2500k @ 4.5 coupled with SLI 680's and saw the same kind of thing. As did Gregster.
 
Just saw what you typed... Isn't that like pretty bad? 98/99 usage is much better right? :) It's what I had on my SLI 480 with my system @ 4Ghz.

If you have 70 - 80% doesn't that mean your system can't give your GTX's as much as they need?

Bf3 is badly optimised also, for both manufacturers. I could have 100fps in one area at 99% usage and then swing around and have 100fps but at 80% usage, with the cpu never going over 75%. It has a lot to do with the amount of people on the map also, for whatever reason.
 
Bf3 is badly optimised also, for both manufacturers.

98-99% usage 6950>70 CrossFire, it was always good.

98-99% usage on 7970/7950 CrossFire with a [email protected], using the 12.9 betas, I can't comment on previous drivers though.

Please don't take this as a dig/slant at anyone/thing, simply being informative for the Nvidia Sli users.:).
 
Last edited:
Wow, ok didn't know even newer i5's were that crippled! Thanks for the information :)
 
98-99% usage 6950>70 CrossFire, it was always good.

98-99% usage on 7970/7950 CrossFire with a [email protected], using the 12.9 betas, I can't comment on previous drivers though.

Please don't take this as a dig/slant at anyone/thing, simply being informative for the Nvidia Sli users.:).

I'm using the 12.9's and I only get that usage in SP and Co-op or on a map with less than 20 players. Any more than that and utilisation craps itself. Are you running the cards at x8/x8 or x16/x16? Also what's your net speeds?
 
Wow, ok didn't know even newer i5's were that crippled! Thanks for the information :)

Wouldn't go that far... it's down to having 4 threads vs 8. BF3 utilises them all and two decent overclocked top end cards are limited by 4 threads in my experience.
 
I'm using the 12.9's and I only get that usage in SP and Co-op or on a map with less than 20 players. Any more than that and utilisation craps itself. Are you running the cards at x8/x8 or x16/x16? Also what's your net speeds?

Long story short, I only had a brief affair with it for about a week, both on 8x8/1100MHz/16GB system ram/OS 128gbM4/BF3 running from 120GB Force 3, mostly on AK too@~10 meg.

It was complete overkill@1080p I may add.
 
Long story short, I only had a brief affair with it for about a week, both on 8x8/1100MHz/16GB system ram/OS 128gbM4/BF3 running from 120GB Force 3, mostly on AK too@~10 meg.

It was complete overkill@1080p I may add.

Are you not using 120hz? The only difference between me and you is net speed, i'm on 1.5/256kb. I'm getting perfect scaling now on 8v8. Anymore and it will drop.
 
It cannot be cpu bottle necking it, cpu usage is all ways around 70%, if it was a cpu problem as in bottle necked, I would expect it to be 100% used, it's not cpu, it's either DICE has messed up or Nvidia have still messed up, take your pick.

Just saw what you typed... Isn't that like pretty bad? 98/99 usage is much better right? :) It's what I had on my SLI 480 with my system @ 4Ghz.

If you have 70 - 80% doesn't that mean your system can't give your GTX's as much as they need?
When exactly are we talking about? now or from what you can remember? as I have heard last game patch screwed up SLI for lots of people, have you checked since last game patch?
 
Last edited:
It cannot be cpu bottle necking it, cpu usage is all ways around 70%, if it was a cpu problem as in bottle necked, I would expect it to be 100% used, it's not cpu, it's either DICE has messed up or Nvidia have still messed up, take your pick.

A CPU bottleneck isn't as simple as that. On my i5 2500k I would see usage between 60-70% on all cores. On my i7 3770k I see the same kind of usage across all 8 threads and GPU usage moved from 80-85% to 95-99%.

It was the same install of Windows and drivers as well. I just took the old CPU out and plopped the new one right in :).

Edit: the only thing I'm not sure on is the degree of bottleneck that will be there on SLI 4** series.
 
Last edited:
A CPU bottleneck isn't as simple as that. On my i5 2500k I would see usage between 60-70% on all cores. On my i7 3770k I see the same kind of usage across all 8 threads and GPU usage moved from 80-85% to 95-99%.

It was the same install of Windows and drivers as well. I just took the old CPU out and plopped the new one right in :).

Edit: the only thing I'm not sure on is the degree of bottleneck that will be there on SLI 4** series.

I don't believe for 1 minute it can be a cpu bottleneck with a 3570k clocked at 4.5ghz with 2 480, what are people using with sli 680's to not get a bottleneck? not all of them are running 3770ks I'm pretty sure of that, if anything most gamers never buy the HT version of a cpu.
 
I don't believe for 1 minute it can be a cpu bottleneck with a 3570k clocked at 4.5ghz with 2 480, what are people using with sli 680's to not get a bottleneck? not all of them are running 3770ks I'm pretty sure of that, if anything most gamers never buy the HT version of a cpu.

Well you don't have to believe it but it's true :D. BF3 MP64 is quite heavily CPU bound actually for some reason. If they aren't running i7's they'll be ever so slightly bottlenecked.

How else would you explain my jump in 10-15% in GPU usage on the same system? Gregster found the same thing when he moved to X79 although I believe his usage readings are a little more over the place.

An i5 is a nice chip, however if you want the utter max from your system and you play BF3 a lot then an i7 is better to get the max GPU usage. Some people are happy to take the 85% scaling with SLI as oppose to chugging more money on a CPU.

As I said, I am somewhat speaking from experience and I was perfectly happy with my set up with an i5 2500k @ 4.5 and SLI 680's. I then sold my CPU and motherboard and upgraded to Z77/Ivy for less than £100. The bump from i5 to i7 and PCI-E 2.0 and 3.0 was about 3-5% in BF3. Not worth it on its own but the reason was the releasing of the slight bottleneck and then minuscule bump from PCI-E 3.0.
 
Well you don't have to believe it but it's true :D. BF3 MP64 is quite heavily CPU bound actually for some reason. If they aren't running i7's they'll be ever so slightly bottlenecked.

How else would you explain my jump in 10-15% in GPU usage on the same system? Gregster found the same thing when he moved to X79 although I believe his usage readings are a little more over the place.

An i5 is a nice chip, however if you want the utter max from your system and you play BF3 a lot then an i7 is better to get the max GPU usage. Some people are happy to take the 85% scaling with SLI as oppose to chugging more money on a CPU.

As I said, I am somewhat speaking from experience and I was perfectly happy with my set up with an i5 2500k @ 4.5 and SLI 680's. I then sold my CPU and motherboard and upgraded to Z77/Ivy for less than £100. The bump from i5 to i7 and PCI-E 2.0 and 3.0 was about 3-5% in BF3. Not worth it on its own but the reason was the releasing of the slight bottleneck and then minuscule bump from PCI-E 3.0.

Well, this is something I found

http://static.techspot.com/articles-info/458/bench/CPU_02.png

To me it just looks like BF3 is just bad with SLI, the only thing that holds it back is a dual core cpu, either that or I got to sell my 3570k and get a 3770k:eek:
 
Last edited:
Same as the last lot according to release notes, but are on a new driver branch.

Some people have noticed reduced boost clocks. Boooo?
Made me cautious to update but my boost clock remains untouched by this driver. Asus TOP 670 @ 1306mhz

It would be nice to identify any similarities in the hardware of those who had their boosts reduced.
 
As the 3D vision side of these been fixed with these drivers, all the drivers since 296.10 have had the 3D vision stop working after a reboot or so, the 3D emitter would not enable in games.
 
Well you don't have to believe it but it's true :D. BF3 MP64 is quite heavily CPU bound actually for some reason. If they aren't running i7's they'll be ever so slightly bottlenecked.

How else would you explain my jump in 10-15% in GPU usage on the same system? Gregster found the same thing when he moved to X79 although I believe his usage readings are a little more over the place.

An i5 is a nice chip, however if you want the utter max from your system and you play BF3 a lot then an i7 is better to get the max GPU usage. Some people are happy to take the 85% scaling with SLI as oppose to chugging more money on a CPU.

As I said, I am somewhat speaking from experience and I was perfectly happy with my set up with an i5 2500k @ 4.5 and SLI 680's. I then sold my CPU and motherboard and upgraded to Z77/Ivy for less than £100. The bump from i5 to i7 and PCI-E 2.0 and 3.0 was about 3-5% in BF3. Not worth it on its own but the reason was the releasing of the slight bottleneck and then minuscule bump from PCI-E 3.0.

I'm with you on this Rusty, I found the same and upgraded on your earlier findings......
 
Back
Top Bottom