GPU bottleneck?

[Waco]

I'd shy away from doing a fakeRAID 5 - your write speeds will SUCK and only load down your CPU even more.

[Shingy]

Decided to make a new thread in the Drives and Storage section

[FlawleZ]

BF Bad Company 2 is heavily CPU dependent. Its one of the few games that likes more cores. Probably one of the most multicore dependent games out currently. Slap a quad in there and you'll see your frames jump up considerably.

[El_Capitan]

BF Bad Company 2 is heavily CPU dependent. Its one of the few games that likes more cores. Probably one of the most multicore dependent games out currently. Slap a quad in there and you'll see your frames jump up considerably.

Where did you get that info? The benchmark test that Tom's Hardware did suggests otherwise: http://www.tomshardware.com/reviews/game-performance-bottleneck,2737-7.html

[RogerDeath]

Where did you get that info? The benchmark test that Tom's Hardware did suggests otherwise: http://www.tomshardware.com/reviews/game-performance-bottleneck,2737-7.html

Of course not with an i5 would there be an issue, but I see my poor little 9950+ constantly being under 80+% load when playing BC2. It depends on the architecture more then clock-speed nowdays...

[El_Capitan]

Of course not with an i5 would there be an issue, but I see my poor little 9950+ constantly being under 80+% load when playing BC2. It depends on the architecture more then clock-speed nowdays...

True. Their tests were mainly done by disabling cores. What's your 9950+ overclocked to? What other processes do you run? Another thing with benchmarks, they're not typical of normal users. Most people have applications running in the background, and that puts more stress on less core CPU's than more core CPU's.

[FlawleZ]

Where did you get that info? The benchmark test that Tom's Hardware did suggests otherwise: http://www.tomshardware.com/reviews/game-performance-bottleneck,2737-7.html

 

There are plenty of sources that prove more cores = better for Bad Company 2. Just look at how the load is distributed across an i7 980x:

 

 

Here's some results from a member on overclock.net when compiling a dual core vs quad core in gaming. This was back when BC2 just released in BETA form which is even more of a testament to how well it likes more cores.

 

Battlefield Bad Company 2 BETA

This was just released a few days ago and dispite some server issues' date=' has been filled with players non stop.

This game takes after Bad Company that was a console game which was very successful.

To get these results I ran a good 30 minute game online. I wanted to get a long game in to insure a good average fps. I used FRAPS to record my results.

Tests run in 1920x1080 4xAA HBAO OFF VSync OFF

Dual Core

[img']http://img684.imageshack.us/img684/6254/bc2dual.jpg[/img]

 

Quad Core

 

Well thats a 2:1 performance ratio. Something tells me this game supports quad cores.

[RogerDeath]

True. Their tests were mainly done by disabling cores. What's your 9950+ overclocked to? What other processes do you run? Another thing with benchmarks, they're not typical of normal users. Most people have applications running in the background, and that puts more stress on less core CPU's than more core CPU's.

I usually only have TS, Steam, and my G35 program open in the background (pretty much at all times). The issue is just that my CPU is old compared to today's standards (go figure that a week or so after I got it AMD released the 920 on me).

[El_Capitan]

There are plenty of sources that prove more cores = better for Bad Company 2. Just look at how the load is distributed across an i7 980x:

 

Here's some results from a member on overclock.net when compiling a dual core vs quad core in gaming. This was back when BC2 just released in BETA form which is even more of a testament to how well it likes more cores.

I'm not sure you can get much merit of a benchmark from a Beta version of a game. Most Beta's aren't even fully stable, much less optimized for performance. I'm not saying there isn't a bottleneck, but from the benchmark that Tom's Hardware did (which was actually a decent test set-up showing the configurations, and not testing the Beta version of the game with an unknown configuration set-up), it doesn't seem to be that big of an issue. Still, there are other variables that come to play that they didn't pick up on, like the architecture difference and background tasks.

[El_Capitan]

I usually only have TS, Steam, and my G35 program open in the background (pretty much at all times). The issue is just that my CPU is old compared to today's standards (go figure that a week or so after I got it AMD released the 920 on me).

Yeah, having those programs open in the background definitely makes a bigger difference in a dual core than a quad core. If Tom's Hardware compared an older architecture dual core to a newer dual core, that would have been good to see.

[FlawleZ]

I'm not sure you can get much merit of a benchmark from a Beta version of a game. Most Beta's aren't even fully stable, much less optimized for performance.

 

Say what you will about the Beta, the facts are BC2 is VERY well multithreaded. Quite possibly the heaviest CPU reliant game we've seen. A simple google search pulls many results of people talking about and experiencing CPU and GPU bottlenecking.

[IVIYTH0S]

Say what you will about the Beta, the facts are BC2 is VERY well multithreaded. Quite possibly the heaviest CPU reliant game we've seen. A simple google search pulls many results of people talking about and experiencing CPU and GPU bottlenecking.

It'd make sense to be, for the Frostbite engine has advanced physics that aren't graphics card accelerated