• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Bf3 and cpu scaling, important read

B

butterfliesrpretty

2[H]4U
Joined
Mar 15, 2002
Messages
2,482
Well I posted this in another thread, but I felt it needed its own because we literally had an entire thread of [H] members talking out their arse. Bf3 is not very cpu dependent at all. Way less than bc2 and most newer games for some reason. Having a good gpu is important though obviously. In fact an i3 dual core will perform virtually the same as a 2600k. Here's your evidence. Do not give advice when you haven't seen the benches.

http://www.techspot.com/review/458-battlefield-3-performance/page7.html
 
We used Fraps to measure frame rates during a minute of gameplay from the single-player mission Go Hunting.
Click to expand...

Yup this tells the entire story about multiplayer alright

Did you even read the thing you linked?
 
socK said:
Yup this tells the entire story about multiplayer alright

Did you even read the thing you linked?
Click to expand...

The ones that were done during the beta which only had multiplayer showed the same thing, from multiple sites. Or should we all just make something up?
 
TwistedAegis said:
60 seconds of gameplay...from single player. What is this, I don't even...
Click to expand...

So do your own tests and prove me wrong, why are you defending people making things up, that's my point, and again multiplayer tests were shown during the beta (albeit it is a different build) and showed the same thing.
 
I'm not trying to prove anything. However bad data can be even worse than no data at all, and putting up a review that has such a terribly flawed foundation doesn't help anything.
 
Furthermore you're all basing the idea that multiplayer will show you entirely different results from one thread that an [H] member made about bc2 that showed a q9550 getting whooped by an i7. I've never seen this duplicated anywhere, we don't know how static or dynamic this test really was. A game like this with many enemies on the screen plus the extra load ai puts on the cpu should show similar results in single player and multiplayer. However this is a mute point because the tests were done again by multiple sites that showed the same damn thing in multiplayer during the beta.

We're all wrong from time to time, having a knee jerk reaction when you're clearly have no evidence backing up what you're saying doesn't make you look that intelligent.
 
I think the OP means well. :)


The thrust of the thread is: Bring evidence to the table. I hope that's what I see here because this is an interesting subject.
 
I'm not trying to say anything, actually, other than that this test is flawed and not likely to be representative. You say it is done by multiple sites, but I only see this one showing up. I'm not claiming to know what BF3's CPU usage is. And does BF3 even use AI in the single player? For the most part it seems to be a few bad guys following predefined paths and shooting at you. I love the game, but the "AI" is incredibly basic.
 
IMO, to really stress the CPU with this game, you'd have to run a 64 player map with everyone destroying everything in sight with vehicle's weapon and RPGs putting the physics to work.

But that's not to say other testing method would be less useful. In fact I think its good to perform these CPU test by different means, multiplayer and single player, and then with all the data, we could observe to see if there's a common trend in how the performance differs with different CPU.
 
Dude, I've done my own tests in multiplayer and noticed a HUGE difference...


64 person Caspian Border
1600x900@120hz
ULTRA Settings
Motion Blur OFF
SSAO

Before:
Q9650 @ 4.275ghz on a Asus P45 mobo.
GTX580 @ 1000mhz
Lowest the fps got was in the 50's

After:
2600k @ 4.9ghz
gtx580 @ 1000mhz
Lowest the fps got was in the 80's

The difference was night and day! ~30 fps is HUGE!


Now don't get me wrong, the Q9650 is still a great CPU, it just can't handle all the action that goes on in a full person large outdoor map.

The differences were not as drastic inside the subway on operation metro, but they were still noticeable.
 
  • Like
Reactions: SoFGR
like this
With the rig in my sig, I watch the task manager and temps on my second monitor while gaming. My CPU load hovers around 40% across all 8 virtual cores. The game defaults everything to ultra, but I have all settings turned to high, due to some release day artifacting and lag issues. I don't run fraps, so I can't tell you what my frames are, but to me they are smooth as butter, specially after the new AMD drivers that came out yesterday.
 
I also watch my usage on a 2nd screen.

This SS below is just a 32person server, but on 64 person servers some cores reach the 100% mark on my 2600k@4.9ghz with HT off...

With HT on CPU usage was around 35-40%.

I get the exact same fps, if anything like 3-5 more fps with it off...

5201settings.png
 
ob1 said:
With the rig in my sig, I watch the task manager and temps on my second monitor while gaming. My CPU load hovers around 40% across all 8 virtual cores. The game defaults everything to ultra, but I have all settings turned to high, due to some release day artifacting and lag issues. I don't run fraps, so I can't tell you what my frames are, but to me they are smooth as butter, specially after the new AMD drivers that came out yesterday.
Click to expand...

If you don't want to use fraps, you can check FPS in game by typing "Render.DrawFps 1" in the console.
 
Caspian is a mother heavy load testing map. I agree that the Q series is not as capable and you should change if you want the smoothest game play. My 2500K at 4.4 and 480 gtx soc runs the game nicely with a min fps around 80 at medium/high settings and SSAO w/ 4XMSAA at 1280X1024. Old crt monitor.

I upgraded from a Q9450 back in BC2 and there was a huge boost. While BF3 is not as cpu dependent, it certainly does benefit going from the Q to the I7. My friend had the highest end Q series cpu clocked to the hilt and a 6990. He got a 2600K on my advice and said the game runs much smoother now.

It isn't a placebo. There is a significant improvement in min fps and thus game play.
 
Last edited:
Spaceman,

I know you have probably gotten use to 1280x1024, but you should use 1280x960 for the aspect ratio to be correct and proportionate when running fullscreen.

CRT's are a 4:3 aspect ratio...

1280x1024 is 5:4 (things are too squished together on your CRT)
1280x960 is 4:3

A cool trick you can do with CRT's is to change your windows resolution on your CRT to 1360x768@85hz or 1600x900@85hz. You can go up to 120hz if you have a higher grade CRT. Then adjust the screen so you have black bars on the top and bottom of the screen. Then change your windows desktop rez to where you had it and load up BF3 and change it to 1600x900@85hz. This will give you a much greater FOV and you will be running a "modern" resolution. You will enjoy 1600x900@85hz on your CRT, let me know how it goes or if you have any questions.
 
Yup. Will try it tonight. Might be why my shot is weirdly off with this game. The disply says that 1280x1024 is the monitors native resolution but idk about that.
 
spaceman said:
Yup. Will try it tonight. Might be why my shot is weirdly off with this game. The disply says that 1280x1024 is the monitors native resolution but idk about that.
Click to expand...
Ya sometimes win7 does that with CRT's. Just remember when you change it to 1600x900 or 1360x768 to max out that refresh rate then adjust the height of the screen all the way down using your CRT's OSD adjustments. It should give you a perfect 16:9 aspect ratio.
 
With my settings on High, SSAO on, Post AA to high and deferred off + vsync: it sometimes drops down to 45fps on Caspian Border.
 
placebo is a helluva drug

lets see some hard evidence, I provided mine. Also feel free to google the beta multiplayer cpu benchmarks that are saying the same thing. I'm not saying all cpus will perform the same, but all of them from this and the previous generation are going to give you damn near the same experience, but I really don't see how a dynamic game experience can be compared in the first place with all the variables out there especially in a game like this. Maybe if it was tested over and over again.
 
Last edited:
hahaha, butterflies you crack me up. Both snappshoota and I have similar results about 30fps gain.
 
butterflies, do you think they're just lying or something with the evidence they've provided you here at [H]? Your "evidence" consists of 60 seconds worth of ONE gameplay scenario in BF3. A 'benchmark' (if it can even be called that) as simple and limited as that means jack SHIT, and we all know it.

Deal with it.
 
stop!theradio said:
butterflies, do you think they're just lying or something with the evidence they've provided you here at [H]? Your "evidence" consists of 60 seconds worth of ONE gameplay scenario in BF3. A 'benchmark' (if it can even be called that) as simple and limited as that means jack SHIT, and we all know it.

Deal with it.
Click to expand...

the vast majority of benchmarks are done like that, you act like this is out of the blue. For all we know we have fanboys trying to justify their purchase by pulling numbers out there ass. I'm not saying thats the case but it is a possibility. Now if you have some youtube demos showing otherwise or a known tech site that ran a multiplayer benchmark MULTIPLE times (at least 5) on the same server with the same amount of people with the same settings, then you might have a point. Otherwise I'm not buying it, and you can deal with it.
 
Actually to keep my setup running at 120+FPS my 5.3 Ghz 2700K is running 80%+ utilization. That is some pretty high utilization for a game.
 
You must log in or register to reply here.
Back
Top