Bad Company 2 Beta Benchmark

[AzN]

http://translate.google.com/transla...field-bad-company-2-beta-gpu.html&sl=ru&tl=en

Ran on a Q9650 @ 4ghz.

From what I gather the game is less than efficient with CPU. CPU usage for dual core is not 100%. More like 70-80%. While Quad is around 50% or so usage.

 

[Qbah]

The HD4890 is getting stomped by the GTX275 - ouch! Let alone a GTX285... HD58xx in the lead, obviously The HD5770s are hurting bad too, HD4870 is quite a bit faster. Looks like you need a HD4870 or faster for a decent experience imo. Or drop the AA.

 

[AzN]

You are right the mainstream cards are being hurt with 8xAA usage but I doubt if you even drop it to 4xAA the performance difference would be closer compared to 4870/4890.

57x0 is more like an overclocked 4850. Consumers just need to accept it instead of trying to compare it to 4870/4890.

Anyways my system ran it less than stellar with all graphic options enabled.

I thought it was my CPU but I ran the game @ lowest settings and average about 60fps and don't dip below 30fps. When I had HIGHEST settings without HBAO my frame rates plummeted into 35-40fps average and dipping into low 20's. I was running 1080P and no AA BTW.

 

[Borealis7]

the 8800GTX is really showing its age when the 4770 and 5750 equal it.
other then some minor irregularities, this graph makes sense, not much surprise here, except if you didnt know a 4860 card existed...

 

[AzN]

Yeah those 4860 are great deal. For the $$$ best performing card in mainstream card price range. I would get that over 5770 or 5750. The potential for those cards rival that of 4870/4890.

 

[vj8usa]

the 8800GTX is really showing its age when the 4770 and 5750 equal it.

Not really - the 8800GTX has always been at roughly the same level as the 4850. The 4770 and 5750 are around there as well, so no surprise there.

Too bad they didn't do any tests with less/no AA - I'm hoping it'll stay at a solid 30+FPS without AA on my 4850.

 

[MrK6]

Something is wrong with this game, either with the game itself of AMD's drivers (which would be kind of odd, considering they're partners for it). I've been playing the beta maxed @ 2560x1600 with 4xAA and I get on average 40-45FPS, which, given the graphics quality, I think is unacceptable. Hopefully things will improve over the next month.

 

[Tab]

I'd be more interested in a CPU test, personally. How much of a difference between AMD X2, AMD IIs and various iterations of Intel Dualcores...

 

[nemesismk2]

I'd be more interested in a CPU test, personally. How much of a difference between AMD X2, AMD IIs and various iterations of Intel Dualcores...

That would of been more interesting for me as well.

 

[solofly]

I'm glad to see my rig can handle it...(as I'm looking forward to this game)

 

[AzN]

http://www.pcgameshardware.com/scre...thread-battlefield-bad-company-2-bc2-beta.jpg

There's your CPU usage with dual core. I don't even think this game is optimized for dual core honestly.

 

[v8envy]

Well, this should shut up people yammering about a 4890 being enough for any game this year. Even a 5870 struggles at 1920x1080.

If this game only uses up 70% of a dual core or 50% of a quad while showing nearly perfect GPU scaling we can rule out CPU being a bottleneck past a certain level. And that level looks fairly low. I'll guess a stock clocked i3 530 would be enough.

 

[hans030390]

Well, this should shut up people yammering about a 4890 being enough for any game this year. Even a 5870 struggles at 1920x1080.

No one needs that much AF/AA.

 

[AzN]

No one needs that much AF/AA.

I played the game with system below. 1080P no AA 4xAF. Playable definitely but not smooth at all. Many dips to low 20's. Hovering around 35fsp.

 

[scooterlibby]

Too lazy to get the link, but the devs have tweeted that they are working on improving dual core performance and that the final edition will have better textures and a more efficient HBOA filter. My system's no slouch, but it struggles with the beta at the current moment.

I'm optimistic about the final edition and patches. Also, hoping to get a high powered DX11 part from either the green or red team. Just holding out for performance numbers from Fermi.

Good info on this game over here http://forums.anandtech.com/showthread.php?t=2020118

 

[Phil1977]

Bad Company 2 really likes to run on a quadcore. In the australian forums we have done lot's of testing...

Even highly overclocked Wolfdale dual cores struggle with this game.

E.g. going from a AMD Phenom 550 to a 955 almost doubles your frames.

I just upgraded my system because the game was basically unplayable on my Athlon II 250. I have a i7 860 now and the game doesn't slow down when more players are on the server.

On the Athlon II 250 is was fine up to 8 players. But once more came online it became so slow it wasn't fun anymore...

Dice mentioned in many interviews that they are going multi core. All their audio processing is done via software. No more EAX and DirectSound3D. They say that the audio processing runs on a single thread.

CPU utilization doesn't reach 100%. On my AMD dual core it was sitting around 70% on both cores. On a quadcore CPU utilization is lower, but it will use all four cores.

Some links to the BC2 discussion in Australia.

Here my CPU loads for the old AMD dual core: http://forums.overclockers.com.au/showpost.php?p=11398913&postcount=1818

This guy did a lot more testing. Comparing Dual against Quad. Very interesting: http://forums.overclockers.com.au/showpost.php?p=11398817&postcount=1809

I am sure Dice will tweak the engine.

This game uses a lot of Physics effects in the game. Most buildings are destructible and the soundstage is really intense. There is a lot more going on than in most single player games.

At first the graphics looks very simple, but once you see all the stuff that happens around you it makes sense that this is a very demanding game. The first few hours I played I got headaches because there was so much going on that it was hard to focus on the game.

Some tips we found to improve performance:

- Force the game to DX9 mode (in your documents/BFBC2 BETA folder there will be text file with config settings
- Disable Bloom in the config file
- Disable HBOA
- Play on servers with fewer players

 

[AzN]

Dice mentioned in many interviews that they are going multi core. All their audio processing is done via software. No more EAX and DirectSound3D. They say that the audio processing runs on a single thread.

CPU utilization doesn't reach 100%. On my AMD dual core it was sitting around 70% on both cores. On a quadcore CPU utilization is lower, but it will use all four cores.

The game isn't multithreaded at least not the this beta version.

Developers claim lot of things but it doesn't necessarily mean it's true.

Your operating system is spreading the cores not the game.

On a side note. I played the game max detail including HBAO but lowering resolution to 720P. Game was much more smoother. Averaging 50fps or so.

 

[toyota]

The game isn't multithreaded at least not the this beta version.

Developers claim lot of things but it doesn't necessarily mean it's true.

Your operating system is spreading the cores not the game.

come on now, a game doesnt have to use 100% of your dual core to be considered multithreaded.

 

[Phil1977]

Your operating system is spreading the cores not the game.

Unfortunatly the OS can't "split up" a single thread across more cores. If that was the case than SuperPi or 3D Mark 2001 would benefit from more cores...

Anyway I won't bother to convince you on this one. You have your views, I have mine

The i7 is awesome for testing as you can just disable cores in the BIOS. And on my machine BF BC2 gets slower the fewer cores I give it.

 

[Borealis7]

come on now, a game doesnt have to use 100% of your dual core to be considered multithreaded.

but its a good start

 

[Phil1977]

but its a good start

I believe 100% cpu usage in games is impossible, unless all the graphics would be rendered via the CPU like in old DOS games. At least I haven't seen 100% usage in any game I can remember. I always have ATI Tray Tools overlay running. It shows the CPU usage along with other probe readings.

But the video card does the rendering now. And the CPU has to wait for the GPU at times. Also some threads need to wait for other threads to finish and sync with each other.

Maybe in a chess game where the graphics are minimal?

 

[AzN]

come on now, a game doesnt have to use 100% of your dual core to be considered multithreaded.

Unoptimized game maybe but can you consider these games actually multithreaded? So now a single core game that spreads the core is dual core optimized all of a sudden?

 

[AzN]

Unfortunatly the OS can't "split up" a single thread across more cores. If that was the case than SuperPi or 3D Mark 2001 would benefit from more cores...

Anyway I won't bother to convince you on this one. You have your views, I have mine

The i7 is awesome for testing as you can just disable cores in the BIOS. And on my machine BF BC2 gets slower the fewer cores I give it.

Vista and Windows 7 "does" split up a single thread game across more cores. Just because it's being spread between the cores does not mean faster performance on a single thread application.

edit:
I just saw the CPU benches you linked to. I replied to you before you edited your post with those links. Interesting that at low resolution a quad can get 80fps average... I rather see some proper benchmarks between CPU so we can clear this up.

 

[schneiderguy]

The game isn't multithreaded at least not the this beta version.

Developers claim lot of things but it doesn't necessarily mean it's true.

Your operating system is spreading the cores not the game.

On a side note. I played the game max detail including HBAO but lowering resolution to 720P. Game was much more smoother. Averaging 50fps or so.

The game gets 50%+ usage on my quad core. Unless Windows is having the same thread run concurrently on two of my cores just "for the lulz" then the game is multithreaded

 

[AzN]

The game gets 50%+ usage on my quad core. Unless Windows is having the same thread run concurrently on two of my cores just "for the lulz" then the game is multithreaded

So what you are saying if the game runs on all your cores it's multithreaded?