World In Conflict: Performance Review

At 1600x1200 things already change and shift towards the upper class of mid-range and more expensive graphics cards. Look at that HD Radeon 2900 XT go by the way. A warning here about the 2900 XT though (before you get too excited) there's a bug in the 7.9 Cats drivers we used and we noticed poorer Texture Quality in certain situations.

For those who just look at the graphs.

Also strange that the reviewer kept mentioning the 8800gts 640mb (never 320) yet only reviewed it with the 320mb card. :confused:

Indeed, a GeForce 8800 GTX is pretty much your minimum required graphics card at 1600x1200. For those that noticed, I ran that 8800 GTS 320 MB test three times to be sure including a re-installation of the driver, but the results matched time after time. I'm pretty confident it's a driver bug though.

Well hopefully they do a full review and use the 640mb card so we can see how it performs in DX10 since the 320mb was obviously having problems (driver or memory, dunno). Hopefully the driver bugs with the 2900 will be fixed and so will the 320mb's DX10 problems.
 
this game ran pretty good on my 8800gts 640 mb and my system in the sig i ran it at 1600x1200 with the stuff on High, was the demo and i did the video test and it was around 30-50 fps
 
Has anyone done much testing with quad core? The review says that the game takes advantage of dual core, I wonder how much quad core helps.
 
this game ran pretty good on my 8800gts 640 mb and my system in the sig i ran it at 1600x1200 with the stuff on High, was the demo and i did the video test and it was around 30-50 fps


In DX10 or DX9? With the default 4xAA 8xAF.. or?
 
What he fails to mention is that he benchmark is worst case scenario for the game. Most of the time you don't have 5 arty strikes, 20 units, and a nuke wiping out a town all at once.

It's a good measure of your lowest expected performance, but most of the time you will be well above that.
 
The game is chocking @ 1920 x 1080 on a QUAD + Ultra.

I'm even using HIGH setting (not VERY high).

Mid 10's in large battles with lots of air support.
 
I like how they dont bench the GTS640 even though it beats the 2900XT
 
this game ran pretty good on my 8800gts 640 mb and my system in the sig i ran it at 1600x1200 with the stuff on High, was the demo and i did the video test and it was around 30-50 fps

I though I read something about fixing the 320MB's performance hole in the 163.69 or 163.71 drivers. Gotta find that source.
 
I like how they dont bench the GTS640 even though it beats the 2900XT

Yeah I didn't get why the reviewer kept saying 640mb whenever he referred to the 8800gts but only used the 320mb card :confused:
 
I though I read something about fixing the 320MB's performance hole in the 163.69 or 163.71 drivers. Gotta find that source.

The issue also involves the hotfixes for Vista, also found on Nvidias driver pages. You need all three, IIRC in conjunction with the 163.xx driver family to get rid of the problem. Oh and that Gamespot review is worthless too, no offense. They use different configurations of hardware and settings which tells you nothing except that they all misrepresent the game as performing at sub-40fps levels.
 
The issue also involves the hotfixes for Vista, also found on Nvidias driver pages. You need all three, IIRC in conjunction with the 163.xx driver family to get rid of the problem. Oh and that Gamespot review is worthless too, no offense. They use different configurations of hardware and settings which tells you nothing except that they all misrepresent the game as performing at sub-40fps levels.


I'm using all the Vista 32 bit hotfixes and the 163.69 driver. The game still performs like a donkey in DX10. It will work perfectly for a while then suddenly drop fps making it unplayable. I'd say Nvidia still haven't fixed this issue properly :mad:
 
it'd be awesome for people who say they're running 1600x1200 and experiencing avg fps of 30-50 with 8800 series cards to step up with some screenies of their in-game settings, fraps average during the last seattle mission, etc.

i don't see how it's possible for a single GTS to achieve those numbers without a massive phase-change overclock on both their quad-core CPU and GPU.

but i'm using the 'lowly' 320MB version combined with an 'abysmally performing' e4300, so that must be my problem... ;)


edit: i love the game
 
Running the benchmark is a worst case scenario. Which is not really seen in the SP campaign but can be expected in MP. Specially using the Nuke a few time a round. Because you can customize so many settings it's hard to tell who's telling the truth. Although for me, turning off certain features does very little to increasing performance when running the benchmark program built in. You'll have to enable or disable several of them, if not all of them, to affect performance. What seems to effect performance the most is:
Resolution
AA
AF
Transparency antialiasing
Soft Shadows
if everything else is left the same as either high/very high or medium/high settings.
 

Good find! Interesting article.
It also looks like that High Quality DirectX 9 looks better then High Quality DirectX 10
It also appears that disabling clouds also disables some ground textures.
High Quality Terrain Enabled/Disabled isn't very noticeable unless you are 1 foot from the ground.
Interesting how they note a decrease in performance from 2 GB of ram to 3 GB at 1600x1200 under XP. But it's the same under Vista.

In all this article provides different benching aspects of this game that have not been seen from other reviewers as of yet. One aspect that I liked was the Settings Test.
 
Back
Top