Battlefield: Bad Company 2 GPU Performance In-depth

Ooooo Awesome !!!! judging by this benchmark my new computer with a 5850 can run this game on highest settings :) I've been laning Battle Field 2 with friends for years now, we never made the jump to Bad Company 1 because our computers couldn't handle it and when they did the game was old already, we are however planning to buy bad company 2 and start the madness :)
 
Just putting this out there. The next iteration of the Frostbite engine has already been tested on up to 16 CPU threads and capable of utilizing them fully, Frostbite 1.5 barely touches 4 threads and or cores. So when the next game comes out i7 will be a nice place to be. That being said, running on my ghetto 8800GT SLI setup I play on High settings at 1920x1080 with no AA or AS and get average 45 FPS with some drops when (you know what) starts going sideways. Also worth noting is my 920 @ 4.0 GHz which is making me want to see what the next engine will be like and what kind of performance boost will be delivered.
 
Indeed. 1943 was supposed to be out on PC close to the console version. But it was then delayed several times. Not it available for pre-order on the EA site (http://www.battlefield1943.com/) due to be released "early 2010". Although I'd guess they will let Bad Company 2 have some time on it's own first
 
Quick update guys the results were recorded with HBAO disabled. The game will not let you view the presets so once we had selected the presets we switched to advanced to show the settings. Although the screen shots showed HBAO enabled it was in fact disabled and this is the performance you will see from these graphics cards with it disabled. I ran some tests on the mid-range cards with HBAO enabled and they became anywhere from 5 – 15fps slower.
 
Julio said:
My bad skitzo, I jumped on it too soon and you are right, Battlefield 1943 was released for consoles only last year and there's a pending launch for the PC. The title shouldn't be missed a lot though now with BC2 out --- 1943 seems to be more of the casual shooter kind, at least that's how it made it big on consoles through digital distribution.
Click to expand...

No worries Julio, now no one else should be confused. And there was always a chance you knew some info that I didn't.

And yeah with BC2 now released I think a 1943 release on PC would flop a retail.
 
We have updated our preset screenshots below noting that all our tests we performed with the HBAO (Horizon-Based Ambient Occlusion) setting disabled in high, medium and low.

The results are exactly the same, our preset screenshots were wrong. Enabling this setting results in lower average fps across the board.
 
the ati cards wouldve been running on dx11 whilst the nvidia cards would be running on dx10/9 unless you changed the setting from Auto in the config so hardly a fair review
 
Steve/Julio,
I have read that the percentage of people who use crossfire is about 2% and even less for Triple/quad crossfireX. so i don't suppose it would be possible to throw one in the mix would it?
 
Guest said:
the ati cards wouldve been running on dx11 whilst the nvidia cards would be running on dx10/9 unless you changed the setting from Auto in the config so hardly a fair review
Click to expand...

Thanks for that little gem but we always test games using the latest DirectX version that the cards can support. You have to edit the settings.ini file to change the mode and since most are not going to do this we test with the default.
 
Looks interesting, and I think my hardware can handle it. Too bad i don't enjoy multiplayer. I don't have 6-18 hours a day to dedicate to any game, so i get my @$$ handed to me by teens and college students.
 
In my own IQ tests (on a 5850 and using the .ini edit), I found, in terms of texture and lighting, that DX10 and DX11 were identical, but much better than DX9. Where DX11 really shines is the realistic softened shadows. DX10 and DX9 had the same uniformly-colored, jagged shadows, whereas the DX11 shadows were soft (darker in the middle, lighter on the edges) and round. It was a subtle touch, but still a nice one. After playing in DX11 and discovering the difference, that particular detail became glaring in DX10, especially when those branches are slapping you in the face as you hump through the jungle (shadows from the other leaves).

I was also able to run the game at 8xAA, 16xAF pretty nicely, especially when ATI Overdrive was maxed out.
 
This tests forget one important thing, that the sound engine is somewhat heavy on the CPUs also, DICE themselves have said that if you want to run it with the highest sound quality which is War Tapes you need a Quad i7 setup. So saying that Duo Core will be enough for this game is wrong if you would like to have the total experience with both high end graphics and sound.
 
Guest said:
This tests forget one important thing, that the sound engine is somewhat heavy on the CPUs also, DICE themselves have said that if you want to run it with the highest sound quality which is War Tapes you need a Quad i7 setup. So saying that Duo Core will be enough for this game is wrong if you would like to have the total experience with both high end graphics and sound.
Click to expand...

I assume you are basing this entirely on something written by the developer and have no real evidence to back up these claims?

Honestly if you believe that a Core i7 processor is required for audio then I have some killer swamp land for sale at a bargain price if you are interested.

Again a high-performance dual-core processor is enough to enjoy this game and a quad-core is not necessary. That said a quad-core is certainly preferred BUT NOT NECESSARY!

Finally with War Tapes enabled we saw only a slight difference in performance when testing. With just two cores enabled on our Core i7 920 processor we were unable to max out both cores.

If you have some real evidence that I am wrong please show me as I am more than happy to take a look at it.
 
Hi Steve,

against all your efforts, at the moment you test is sadly useless to reflect the gpu and cpu scaling performance of BBC2.

1: You didnt test the game with the day 1 patch installed.

- You should have known that within the first patch a lot of major performance optimizations are included, which gives you an performance boost up to 30% on many systems.


2: You did the test on the singleplayer campaign.

- But most important for the longtime experience of BBC2 is the multiplayer part. And it looks like the mp is much more performance intensive and cpu and gpu depending than the singleplayer campaign, especially on filled 32 player servers. The results there may differ a lot.

So plz, redo the test under this terms and it your results would really reliable.

regards
 
The article very accurately displays GPU performance and the bench testing was completed before the game/first patch was released but thank you for telling me what I should have known. That said having re-tested with the first patch none of the graphics cards displayed any tangible gains. If there were 30% gains to be had we would re-test but there simply isn’t.

I cannot accurately test the game in a multiplayer environment, it is simply not possible without timedemos.
 
Steve and et al,

I found your article to be straight forward, very informative and well balanced. I am curious though, you mentioned on page three of the article that,

"However Crossfire was full of graphical glitches and although we did not experience any crashing, there were a number of rendering issues that still need to be solved"

Have you done any additional testing since the patch that would indicate all of the issues have been corrected? I am still experiencing the the flashing and flickering and the only answer I get from either EA or ATI is that the issue is fixed. The only way I stopped it was to turn off the AI in the CCC. Just curious.

Thanks and keep up the good work.
 
ummm onli 1 thing, why you dont make a bench with a sli card set, i know lot of people like ati(for me is the same ati or nvidia, perofrmance acordi to my need thats what i have instaled, have both cards sets for diferent propouse).

think a sli comparsion can be more fair for both companies.

the game look so cool that why also, hope i can play it in a few.

have fun
 
Very good article. I think the graphics of bad company 2 are amazing, and run particularly well on most systems. I have an 8800gt and run it on high with 1xaa, 1xaf 1920 x 1200 res at about 40fps average. Excellent game.
 
these people should make games such that every common man can afford to play it! by increasing the minimum requirements they are hurting their own pockets!
 
Good article but really applicable for single player only. MP is very very much more demanding of CPU with all the physics calculations. Dual core Athlons especially have a terrible time of it, no matter what video card.
 
Hi all,

rumbel the Test:

Go into the Battlefield Bad Company 2 -Gamesettings.ini-
and Test with different Render Ahead Settings.

Default is 2

But you can use 0 - 4 and have different Feeling and CPU/GPU use.
I test on a second Monitor with the Task Manager and Fraps and Everest.

Try it and you will see a different.

Rigger
 
i am just wondering in what setting the console version is running this game cause their GPU is nowhere near GTX260 pr 4780.
 
I wish you would have actually tested some dual core processors rather than just disabling cores. I don't see how you can say that CPU is taken out of the equation when I've talked to many people that are running Geforce 8800GTS cards and while dual core users are just barely playable, quad core users are reporting decent performance.
 
You must log in or register to reply here.

Similar threads

Daniel Sims
Intel XeSS 1.3 improves performance with new presets, intros DLAA alternative
Replies
12
Views
138
Disiw
D
Jimmy2x
Modern Warfare 2's season 4 update continues to miss the mark
Replies
10
Views
525
Nobina
Nobina
D
AMD's new flagship Radeon RX 7900M mobile GPU comes with 4,608 cores & 16GB VRAM
Replies
12
Views
527
Watzupken
W

Latest posts

Back