BF3 64 player MP CPU benchmarks!

There's no way a Q6600 only gets min fps of 40 more like min fps around high 20's. And hitting frames under 40 in this game is not enjoyable at all especially in high intense areas. Might as well just stand there and die :)
 
rushaoz said:
So this is awesome news. Can you possibly check something for me? Run strike at karkand. With 64 players and keep your eye on fraps. See what your lowest fps is on that map. I'm in the high 20s. Basically I want to know what my gtx 460 is capable of before I upgrade. If going to am i5 2500k @ 4.2 ghz means my gtx 460 wont drop below 45 fps under any circumstance in bf3... I'm in.
Click to expand...



Did you ever get the data you wanted?

If not, you should include some more info. Specifically, what video settings. DuronClocker stated that he is using 2X MSAA. I have a 2500K at 4200MHz and my GTX 460 is at 850MHz. I haven't checked SaK 64 player but before the expansion came out I was checking my fps from time to time and I usually got between 60~80fps with minimums mostly at 50fps and above.

This was with custom settings of mostly high except for no AA, motion blur off and SSAO.
 
jreinlie4 said:
Your probably going to see significant improvement, especially if you OC it.
Click to expand...

*UPDATE* Yep....BF3 and Skyrim run like new games now, getting 80-100fps in ultra settings 1900x1200 with dips in the mid 60s sometimes 6950 1GB Crossfire. So far no stuttering at all. Very happy. :)
 
neokeelo said:
*UPDATE* Yep....BF3 and Skyrim run like new games now, getting 80-100fps in ultra settings 1900x1200 with dips in the mid 60s sometimes 6950 1GB Crossfire. So far no stuttering at all. Very happy. :)
Click to expand...

I was in the same situation, coming from a Q9550 (and x48 chipset) with 2x5870 / 1920x1200, now on 2500k @ stock..

Quite the improvement, about 20-30% in speed and no stuttering at all, more min fps. I'm going to install my Zalman 9700led and OC as soon as my lga1155 retention bracket arrives. :D
 
i would like to see this same test with 1100T @ 4.2GHz and same GHz for fx 8150

i wonder how six REAL cores and eight REAL cores stack up to I5 and I7 @ 4.5GHz but with only 4 REAL cores and 4 Fake Cores (HT)

compare that someone
thanks
 
tdbone1 said:
i would like to see this same test with 1100T @ 4.2GHz and same GHz for fx 8150

i wonder how six REAL cores and eight REAL cores stack up to I5 and I7 @ 4.5GHz but with only 4 REAL cores and 4 Fake Cores (HT)

compare that someone
thanks
Click to expand...

1. Why not just run them all at the same speed?
2. AMD's eight half-cores aren't much more real than Intel's HT cores.

You can claim that there's real circuitry to back up AMD's half-cores, but you'd be glossing over the reason HT provides any boost at all: there's more resources in each of Intel's cores than any single thread can make use of at once. So in reality, HT is backed up by real circuitry too, and it's just accessed virtually.

Further, and sorry to disappoint you, you'll find that the AMD CPUs are going to eat up more power while pushing out lower numbers all around. They're over a generation behind Intel.

Oh, and the Intel CPUs cost less and overclock higher. Go grab a 2500k, sans HT, and run it at 4.8-5.0GHz. Then watch it beat the piss out of every AMD product on the market for less.

(I don't hate AMD BTW, but I am extremely disappointed in them)
 
poj said:
I was in the same situation, coming from a Q9550 (and x48 chipset) with 2x5870 / 1920x1200, now on 2500k @ stock..

Quite the improvement, about 20-30% in speed and no stuttering at all, more min fps. I'm going to install my Zalman 9700led and OC as soon as my lga1155 retention bracket arrives. :D
Click to expand...

I made this jump for BF3 too, based on BF:BC2's CPU usage. The Q9550 was enough for that game, but it was pegged, while my GTX570 was running at 50-70%.
 
I actually want to see this now

1090T/1100T @ 4.2Ghz or above (REAL SIX CORES)
FX 8150 @ 4.2Ghz or above (4 real 4 fake)
intel 6core @ 4.2Ghz or above (REAL SIX and disable all HT)

the reason I am pointing this out is because.

for SP we know a highend dual core is all you need
for MP in BIG maps (64) we definitely know multi-core plays a HUGE difference in FRAME RATE.

jreinlie4 Post #74
its all SP you did your benchmarks on
so it didn't help to much

can someone with a highend setup (six real core cpu) and highend gpu setup do some tests for us on 64 player maps in multiplayer?

here is what I would like to see as far as settings.

enable all cores and disable all HT (if applicable)
overclock to 4.2GHz if possible
1080P with LOW settings
then benchmark FPS (let us know if you are being bottlenecked by GPU as it makes all the difference)
record those FPS somewhere

then repeat same test with ULTRA

then repeat both tests with only FOUR cores enabled and keep HT disabled.

what this will tell us is:

is there actually a frame rate decrease when going from 6 to 4 real cores (which no one has done yet).

thanks
 
it can be ANY SIX CORE CPU as long as it is a "real physical cores" which would also include the cheaper 1055/1065/1090/1100 cpu's from amd

this test should be relatively easy since intel made ALOT of SIX CORE cpu's also

there should be quite a few (i think the real problem is probably the graphics card to NOT bottleneck the tests
 
tdbone1 said:
it can be ANY SIX CORE CPU as long as it is a "real physical cores" which would also include the cheaper 1055/1065/1090/1100 cpu's from amd

this test should be relatively easy since intel made ALOT of SIX CORE cpu's also

there should be quite a few (i think the real problem is probably the graphics card to NOT bottleneck the tests
Click to expand...

The cheapest Intel six core is still $500, and shows almost no advantage in gaming over their $200 quads when overclocked, including in BF3.

If you want to compare six cores versus four cores that's great, but most reviews have shown the difference to be meaningless for gaming, even if the game in question can use the cores.
 
there are NO reviews that show six core vs four core.
why do you think i have responded in this thread.

this is the only thread that is even close to it where they started with 4 cores and went down to 2 cores
MULTIPLAYER TESTING is the thing to remember
only one review i could find and that was of "bf3 beta" almost a year ago before bf3 final was released

anyhow you can get six core intels off ebay for $200 so im not gonna pay attention to what you said about intel for $500

and you can get the 1055/1090/1100t's $200 brand new so im sure they are running anywhere from $120 to $150 on ebay again

this test should easily be done as this is pretty popular hardware.

just gonna take the right person to see it here
 
CPU matters a lot in 64 player games it seems like. Best to setup your browser to run bare without any plugins to get the best performance possible when playing in game.
 
best to see if 64 MP makes a diff in fps going from >4.2ghz 6 core to 4 core
we would like to know just like the ppl with 2 core to 4 core wanted to know
 
tdbone1 said:
best to see if 64 MP makes a diff in fps going from >4.2ghz 6 core to 4 core
we would like to know just like the ppl with 2 core to 4 core wanted to know
Click to expand...

Not that I wouldn't want to know as well.

Also, I can't consider used prices on Ebay, especially with CPUs. You have no idea what you're getting and when it might fail, and if you care at all, you'll get a new one.
 
the intel warranty is covered from person to person so its ok as long as its under warranty

I noticed you keep skipping the amd 1055/1090/1100 chips
those are everywhere
 
anyone want to do these tests?

we have the 4 core vs 2 core comparison how about the 6 core vs 4 core now?
 
I used fraps to track my framerates for 10 minutes. (fraps is actually known to perform slightly worse than MSI Afterburner).

1080p all settings maxed, except I'm not using any of the in game AA. I inject SMAA Anti-Aliasing with sweetFX. It performs slightly worse than the in game "post" (which is some form of FXAA) but gives the same or better level of edge smoothing, and doesn't blur textures like FXAA. I'm also using SweetFX to run the color saturation booster and slightly boosting sharpness. I believe it uses videocard shading power to do that.

so in summary, since i'm using fraps, SMAA, and taking some shading power to boost color saturation and sharpness----my numbers should represent a worst case scenario for my system!

my system is the following:

AMD Phenom II X6 overclocked from 2.8 to 3.7 GHZ
8GB of DDR3 overclocked from 800mhz to 886
my mother board northbridge is overclocked from 2000mhz to 2658mhz. (known to boost AMD peformance).

HIS Radeon 7870 overclocked gpu core from 1000mhz to 1150. GPU memory overclocked to 1325mhz (think stock is 1200mhz). This is a modest overclock and means my 7870 does not approach 7950 performance.


here are the numbers:
bf3frapslog.jpg


The first set is from "sabalan pipeline", physically a smaller map

The second set is from "Operation Riverside", this is a larger map.

both maps had over 58 players at the time of testing. Again, the numbers were benched over a period of ten minutes.



I will post some benchmarks later, with the Notherbridge set to stock speed, to see how this affects BF3.

I may also disable two cores and see how that peforms!
 
Last edited:
chameleoneel said:
I used fraps to track my framerates for 10 minutes. (fraps is actually known to perform slightly worse than MSI Afterburner).

1080p all settings maxed, except I'm not using any of the in game AA. I inject SMAA Anti-Aliasing with sweetFX. It performs slightly worse than the in game "post" (which is some form of FXAA) but gives the same or better level of edge smoothing, and doesn't blur textures like FXAA. I'm also using SweetFX to run the color saturation booster and slightly boosting sharpness. I believe it uses videocard shading power to do that.

so in summary, since i'm using fraps, SMAA, and taking some shading power to boost color saturation and sharpness----my numbers should represent a worst case scenario for my system!

my system is the following:

AMD Phenom II X6 overclocked from 2.8 to 3.7 GHZ
8GB of DDR3 overclocked from 800mhz to 886
my mother board northbridge is overclocked from 2000mhz to 2658mhz. (known to boost AMD peformance).

HIS Radeon 7870 overclocked gpu core from 1000mhz to 1150. GPU memory overclocked to 1325mhz (think stock is 1200mhz). This is a modest overclock and means my 7870 does not approach 7950 performance.


here are the numbers:
bf3frapslog.jpg


The first set is from "sabalan pipeline", physically a smaller map

The second set is from "Operation Riverside", this is a larger map.

both maps had over 58 players at the time of testing. Again, the numbers were benched over a period of ten minutes.



I will post some benchmarks later, with the Notherbridge set to stock speed, to see how this affects BF3.

I may also disable two cores and see how that peforms!
Click to expand...

frapslog2.jpg


This is Kiasar Railrod. All entries seen there are benched over 5 minutes each. From the same map, on the same server, in the same match (yes, even though I rebooted at one point, I was able to enter the same match and get a full 5 minute bench, before the round ended!). For each run, I made sure to get on a dirtbike and wing across the map in third person view, in addition to just playing the game without thinking about anything else.

For the first entry, I tried physically disablling 2 cores in the bios, to make my phenom II x6 like an x4. but for some reason the computer wouldn't post at all. So instead I went to the BF3 process and set the processor affinity to only use 4 cores. So BF3 was running on 4 cores, while everything else (windows, GPU, etc) was running on 6.

for the second entry, I alt tabbed and restored the processor affinity to 6 cores, then returned to the game and benched again.

for the third entry, I completely closed the game and restarted it with all six cores in affinity, just in case messing with the affinity mid game somehow compromised the 6 core performance.

for the fourth entry, I rebooted the machine and de-clocked the northbridge overclock from 2659mhz, to as close as I could get it to stock on the multipliers i'm using, which was 2127mhz.

so in summary:

1. 4 core BF3 affinity
2. alt tabbed and restored to 6 core affinity
3. completely exited and restarted BF3, with default 6 core affinity
4. returned northbridge to near stock clock.

notes:
I used fraps to track my frameratess. (fraps is actually known to perform slightly worse than MSI Afterburner).

1080p all settings maxed, except I'm not using any of the in game AA. I inject SMAA Anti-Aliasing with sweetFX. It performs slightly worse than the in game "post" (which is some form of FXAA) but gives the same or better level of edge smoothing, and doesn't blur textures like FXAA. I'm also using SweetFX to run the color saturation booster and slightly boosting sharpness. I believe it uses videocard shading power to do that.

so in summary, since i'm using fraps, SMAA, and taking some shading power to boost color saturation and sharpness----my numbers should represent a worst case scenario for my system!

my system is the following:

AMD Phenom II X6 overclocked from 2.8 to 3.7 GHZ
8GB of DDR3 overclocked from 800mhz to 886
my mother board northbridge is overclocked from 2000mhz to 2658mhz. (known to boost AMD peformance).

HIS Radeon 7870 overclocked gpu core from 1000mhz to 1150. GPU memory overclocked to 1325mhz (think stock is 1200mhz). This is a modest overclock and means my 7870 does not approach 7950 performance.
Click to expand...

oh and if anyone is curious, my motherboard is a M4A89GTD PRO/USB3, which I bought back in 2010 due to [H]ardocp's very positive review of it.
 
Last edited:
You must log in or register to reply here.
Back
Top