Realtime CPU benchmark for BFME2 owners [Archive]

View Full Version : Realtime CPU benchmark for BFME2 owners

SH64

10-01-06, 09:56 PM

Ok since i got my E6700 i couldnt really tell how powerfull it is because i couldnt see any real benefit in real games since most of them are usually GPU limited. i mean yeah i get higher scores in SuperPI & 3DMark , but thats nothing if theres no significant advantage in real-world games.

anyway long story short i took LOTR : Battle For Middle Earth 2 for a spin with the new Conore CPU to test it againest the 4800+ .. i asked my friend to make a savegame with insane amount of army to see how the processors with cope with their AI & pathfiniding .. etc
FYI BFME2 is very CPU demanding esp when you build a large army & get into battles.
i took the savegame & with little mod i turned it into a CPU benchmark & here are some quick results :
1) notice the test will stress only the CPU , the GPU seems to have no effect ( i made sure of that by running the bench on both 800x600 & 1280x1024 res .. exact same results ).
2) Idnetical settings used & high-end GPUs.

3 systems where used .. 2 of mine & 1 of my friend's .

4400+ @2.3Ghz = 8fps
4800+ @2.4Ghz = 9fps (~12.5%+ over 4400@2.3)
E6700 @2.7Ghz = 14fps ! (55.6% + over 4800+)

Now if there are enough members here interesed on contributing on the benchie i'll be happy to upload the savegame & give instructions on how to do it .. & i'm sure this bench will be very helpfull for overclockers to test o/c advantage in a realtime bench aswell as getting ready for C&C3 which will be using the same engine.
if there isnt enough people interested in this (or dont have the game actually to test) .. well i'll leave it at that in hope that some readers can make use of it or take notes.
:)

EDIT : see post #6 for the savegame & instrcutions.

ArturNOW

10-02-06, 05:57 AM

Is that game multithreaded?

Is it possible to disable one core and benchmark it again?

SH64

10-03-06, 06:52 AM

Is that game multithreaded?

Is it possible to disable one core and benchmark it again?
I dont think it is .. but i can try!

Stoneyguy

10-04-06, 10:33 AM

sure I'll give the game save a try.

I'm running:

6400 @2.44
1 gig of DDR 333 Hyper X at 2 CL
7800 GS oced to 1330 memory and 440 GPU

Redeemed

10-05-06, 06:00 AM

sure I'll give the game save a try.

I'm running:

6400 @2.44
1 gig of DDR 333 Hyper X at 2 CL
7800 GS oced to 1330 memory and 440 GPU

I don't think the game would need to make use of dual-core systems. I'm using an A64 3400 s754 and it runs fine on my rig. And I have everything maxed at 1920x1440. Not sure how it could benefit from a dual-core setup.

SH64

10-05-06, 08:12 AM

Ok heres the save game .. notice that it was taken with patch 1.05 (dunno if its going to be compatible with the new 1.06 patch) :

http://rapidshare.de/files/35572658/SH64_CPU_Benchmark.rar.html

instrcutions on how to bench :
1) Preferable settings : 1280x1024,4xAA,16xAF, Ultra high IQ. (thats why a High-end GPU is recommended to make most accurate scores)
2) load the saved game from skirmish
3) when game is loaded DONT move the mouse cursor to/or leave the scene !! just press "Q" (defualt key to seclect all battalions) & then right click on the ground so the whole selected army will come over to the scene.
4) Now watch your framerate with FRAPS after the right click & write down your most minimum framerate!
thats it! :D .. sorry its a little tricky , but you have to do it this way to get accurate results.

now if your most minimum framerate is in :
- [5-9fps] i.e. single digits then you are in avarage speed
- [10-15fps] then you are in good speed
- [16-20fps] = high speed
- [21+fps] .. you are a GOD , you shouldnt be here talking to us!! :D

& finally i'll stress again : the bench has nothing to do with your GPU if its a high-end .. its all about CPU computation power. you can try the benchmark @800x600 & up to 2048x1536 , you will get the same min framerate.

Stoneyguy

10-05-06, 11:55 AM

SH64

10-05-06, 12:03 PM

10 fps

and it just barely flashed 10 twice. :(

On a side note: I'm new to the fraps program. I set it to record the min/max/average but I can't find it anywhere. The readme says it's in a benchmark subdir but I can't even find that subdir. Do I need the full paid version or am I missing something?
Well 10fps is not bad .. just to show how BFME2 can be a CPU killer!!

BTW : you said I'm running:
6400 @2.44
1 gig of DDR 333 Hyper X at 2 CL
Is that a Core 2 Duo 6400 ?? how come you are using DDR333 with it ?? .. maybe you listed something wrong here ?

as for FRAPS you should find the results in the main dir (ex. C:\Fraps) under tha name FRAPSLOG.TXT

Stoneyguy

10-05-06, 06:06 PM

No I didn't list anything wrong. I'm using the ASrock mobo that allows you to use DDR 266,333,400 along with DDR2 (but not DDR2 800). That's also how I'm able to use the 7800 GS as it's an AGP card.

Can't find the Frapslog.txt either :(
I guess I did something wrong.

fatal1ty

10-06-06, 10:27 PM

Tygerwoody

10-07-06, 09:27 AM

I don't like seeing percentages when the numbers are so low. Like when ATI releases drivers and the release notes claim a 55% increase in game X at 1600 x 1200 with 6xaa / 16x af... what kills me is when 55% = 4 or 5 FPS thats just LAME!!!!!!

I see your excitment...realizing that the game MIGHT be CPU dependant but come on man if you are using a 939 with DDR 1 system and have to fork out $200 for a decent conroe motherboard, another $525 for the e6700 chip and another $400 for decent 2GB of DDR 2 then I sure as hell better see more of an increase than 5 fps...I almost died laughing when you noted it to be 55.6% increase over 4800, I'm sorry but 14 fps still sucks LOL.

4400+ @2.3Ghz = 8fps
4800+ @2.4Ghz = 9fps (~12.5%+ over 4400@2.3)
E6700 @2.7Ghz = 14fps ! (55.6% + over 4800+)

Wow I'm sold on a conroe time to sell my 3Ghz overclocked X2 3800 and blow $1100 to gain 2 or 3 more FPS... W00000t!
:rolleyes:

You are crazy if you dont think that is a noticable improvement. 55% is still 55% no mater how many ways you cut it. And your logic is thoroughly flawed by saying you "blow $1100" on the system. So if I have a 500 dollar video card and I sell it... then buy a 600 dollar video card is that blowing 600 dollars? I highly doubt you would say that. Bottom line is that the conroe system costs about a grand... you a right. However, if you were to build an Athlon system to try to get remotely close to those scores you will be spending ALOT more money on some FX chip(and then if you bring overclocking into the scenario those Conroe chips overclock WAY WAY better). I'd say it was money well spent.

Nice scores btw. ;)

Stoneyguy

10-07-06, 11:10 AM

Well I spent 300 bucks and got an ASrock mobo and a 6400. I kept my memory and my 7800 GS card. Using this setup and a mild overclock (the board is a poor overclocker) of 2.45 from 2.13, I'm able to best a stock FX-62 in most of Sandra's benchmarks. Yes I know it's synthetic benchmarks, but even though to spend only 300 bucks and get performance that is even in the same ballpark as a FX-62 is amazing.

SH64

10-07-06, 01:16 PM

I don't like seeing percentages when the numbers are so low. Like when ATI releases drivers and the release notes claim a 55% increase in game X at 1600 x 1200 with 6xaa / 16x af... what kills me is when 55% = 4 or 5 FPS thats just LAME!!!!!!

I see your excitment...realizing that the game MIGHT be CPU dependant but come on man if you are using a 939 with DDR 1 system and have to fork out $200 for a decent conroe motherboard, another $525 for the e6700 chip and another $400 for decent 2GB of DDR 2 then I sure as hell better see more of an increase than 5 fps...I almost died laughing when you noted it to be 55.6% increase over 4800, I'm sorry but 14 fps still sucks LOL.

4400+ @2.3Ghz = 8fps
4800+ @2.4Ghz = 9fps (~12.5%+ over 4400@2.3)
E6700 @2.7Ghz = 14fps ! (55.6% + over 4800+)

Wow I'm sold on a conroe time to sell my 3Ghz overclocked X2 3800 and blow $1100 to gain 2 or 3 more FPS... W00000t!

I see your point & i'm not here promoting for the Conore , otherwise i'd post this thread in the CPU section .. its all about benchmark results.

2) like what Tygerwoody said : 55% difference is 55% .. sole framerate numbers are not the issue here unless the reader is some kind of a n00b who dont understand simple math.
now if i say - for another game - that a 4800+ gets 25fps & the Conore gets 39fps would that make you or the reader happy ?? because its the same percentage afterall.
what matters though .. at the end of the day the guy with the E6700 had better experiance in BFMEII than the guy with the 4800+ (i.e. higher minimum & avarage framerate).

3) As for upgrading .. well you have to go DDRII sooner or later .. its the future. both AM2 & Intel Core 2 chipsets requires it so at least you're not wasting money by getting those.

4) Those are stock numbers & everyone here knows that the Conore is a hell of an overclocker (& much better than the X2s at that) .. so with a little o/cing you could achieve 15fps+ with no problems.

so let me stress again that from a *technical standpoint* this is a very nice increase in performance for a CPU dependent game.

fatal1ty

10-07-06, 01:39 PM

:rolleyes:

You are crazy if you dont think that is a noticable improvement. 55% is still 55% no mater how many ways you cut it. And your logic is thoroughly flawed by saying you "blow $1100" on the system. So if I have a 500 dollar video card and I sell it... then buy a 600 dollar video card is that blowing 600 dollars? I highly doubt you would say that. Bottom line is that the conroe system costs about a grand... you a right. However, if you were to build an Athlon system to try to get remotely close to those scores you will be spending ALOT more money on some FX chip(and then if you bring overclocking into the scenario those Conroe chips overclock WAY WAY better). I'd say it was money well spent.

Nice scores btw. ;)

My point was if you already own the 4800+ system, which was being compared. If you are building a new system, of course Conroe is the way to go. My argument is that I don't find conroe worth the cost of the upgrade right now for gamers. There is little benefit in most games at high resolutions and a from 9 - 14 FPS at that resolution tells me that a new GPU will gain you a helluvalot more than dumping an existing 4800+ system for a conroe system.

If someone has an existing AMD dual core system - sit on it, buy the fastest GPU you can afford (for gamers anyways). Gaming, at this point, is still GPU limited. Benchmarks at 640 to show which CPU is faster are moot unless you game at 640.

fatal1ty

10-07-06, 01:43 PM

I see your point & i'm not here promoting for the Conore , otherwise i'd post this thread in the CPU section .. its all about benchmark results.

2) like what Tygerwoody said : 55% difference is 55% .. sole framerate numbers are not the issue here unless the reader is some kind of a n00b who dont understand simple math.
now if i say - for another game - that a 4800+ gets 25fps & the Conore gets 39fps would that make you or the reader happy ?? because its the same percentage afterall.
what matters though .. at the end of the day the guy with the E6700 had better experiance in BFMEII than the guy with the 4800+ (i.e. higher minimum & avarage framerate).

3) As for upgrading .. well you have to go DDRII sooner or later .. its the future. both AM2 & Intel Core 2 chipsets requires it so at least you're not wasting money by getting those.

4) Those are stock numbers & everyone here knows that the Conore is a hell of an overclocker (& much better than the X2s at that) .. so with a little o/cing you could achieve 15fps+ with no problems.

so let me stress again that from a *technical standpoint* this is a very nice increase in performance for a CPU dependent game.

I wasn't trying to be a dick, my post above just explains my sentiments a bit more. I just hate seeing people dumping perfectly good AMD systems for conroe when a faster GPU is money better spent still. My current rig is competing with conroe / 7800 GTX SLI / 7900 GTX SLI / Crossfire rigs in 3dmark 06 and 05. My x2 3800 is running at 3 Ghz with my memory running DDR500 2.5-3-6-2. I managed 9300 in 3dmark 06 and 14,600 in 3dmark 05 - with an X2 and a 7950 gx2.

A well tuned X2 system can hold it's own with conroe in high res gaming since the GPU is still the bottleneck.

SH64

10-07-06, 01:52 PM

I wasn't trying to be a dick, my post above just explains my sentiments a bit more. I just hate seeing people dumping perfectly good AMD systems for conroe when a faster GPU is money better spent still.
..
A well tuned X2 system can hold it's own with conroe in high res gaming since the GPU is still the bottleneck.
& i didnt say otherwise ;)
i didnt ask people to dump their AMD systems nor said the Conore pwns the X2 in GPU dependent games. actually i hardly see a difference in GPU dpenedent situations.
thats why i stressed in the title CPU benchmark.

i hope people who read this thread understand that its simply a good realtime benchmark to see how your CPU compares to other ones because seriously i had enough testing with synthetic benchamrks & its time for the real thing.