VR-Zone: Counter-Strike:Source Benchmarked!

[martrox]

These results seem pretty much the average of almost all games. The X800PE is the fastest, and the 6800Ultra and X800Pro are very close, with the 6800GT running a bit behind and the 6800NU running quite a bit behind.

Which just makes DOOM III the really odd game out.

I just wonder what these results have to do with HL2.... probably not much!

 

[jb]

Wow, looks pretty good for playable frame rates on most of the r3xx+ and NV35+ class of cards!!!

 

[trinibwoy]

BTW-Devourer over at EB pointed out to me that they used dx9.0b and that the 6800 needs dx9.0c to enable some of the zippy new features, wouldn't that affect the 6800's numbers? :|

Counterstrike Source uses dx9.0c features? Also, me being a virgin to the half-life/counterstrike scene what exactly is the hype about this game - is it a makeover of the original with HL2 class graphics?

 

[digitalwanderer]

Counterstrike Source uses dx9.0c features?

I honestly don't know, but I felt it worth pointing out. There are also a few other discrepencies in the rig specs. :?

Also, me being a virgin to the half-life/counterstrike scene what exactly is the hype about this game - is it a makeover of the original with HL2 class graphics?

Yup. Just the original Counterstrike using the Source engine, and they're only releasing one map for the beta test period.....yet I'm still checking every 20 minutes or so to see if the beta is active.

(Yeah, I'm one of those saps who is excited as hell about the beta CS:S starting today. I don't think I've ever even played CS, I'm just dying to take a look at the Source engine in action!)

EDITED BITS: The discrepencies and problems noted at Elite Bastards in the review:

TheDon]It gets even more odd:

• MSI nForce 3 Ultra Platinum (Socket 939)

• VIA Hyperion 4in1 4.53

:?

Really the only time the R420 based cards performed differently than their R420 counterparts

R420 performing differently from R420?

 

[jvd]

Counterstrike Source uses dx9.0c features? Also, me being a virgin to the half-life/counterstrike scene what exactly is the hype about this game - is it a makeover of the original with HL2 class graphics?

What do u mean by c features . The source engine supports up to sm 2.0b . I don't believe there is a sm 3.0 path in it .

This is basicly a counterstrike with updated graphics . The textures are about the same as the original just the geometry and a few other things have been upped . . Its basicly for those who have geforce 4- x800xtpes that love counter strike but were running a 5 year old engine with such powerfull cards.

Hl2 will basicly just up the graphics more than this. Much more . It will also support 3Dc .

 

[trinibwoy]

What do u mean by c features . The source engine supports up to sm 2.0b . I don't believe there is a sm 3.0 path in it .

Thanks. I was responding to an assertion above that performance might be improved on 6800's with dx9.0c and I was doubtful since counterstrike is such an old game.

Hl2 will basicly just up the graphics more than this. Much more . It will also support 3Dc .

OK.

 

[trinibwoy]

R420 performing differently from R420?

Probably just a typo. But maybe he was talking about XT vs PRO as they are pretty close unless at high res with AA/AF as his comments indicate. But then all the cards are close at those settings so it's anybody's guess what he meant.

As you can see, with no Anti-Aliasing or Anisotropic Filtering enabled, every card besides the GeForce 6800 standard does very well in Counter-Strike: Source.

This after the 6800 - only card with 128 ram - stays head to head with all other cards till 12x10 noAA and 10x7 AA/AF. What's this reviewer smoking? Would be interesting to see how a vanilla 6800 performs with 256 ram.

Anyhoo...what's up with the jagged edges on the gun in the AA test for the 6800U?

 

[T2k]

does cz use heavy shading like HL2?

No way!

Condition Zero is based on the original Half-life engine.

True. But what CZ has to do with CS:S?

 

[AlNom]

Anyhoo...what's up with the jagged edges on the gun in the AA test for the 6800U?

I noticed that too; I was wondering why the reviewer didn't notice it. It looks pretty bad too.

 

[jvd]

This after the 6800 - only card with 128 ram - stays head to head with all other cards till 12x10 noAA and 10x7 AA/AF. What's this reviewer smoking? Would be interesting to see how a vanilla 6800 performs with 256 ram.

Thats really what hte problem is though. 1027x768 is good with aa and af. But in the source engine its not as good of a buy as other cards.

Wait till the second part comes out and u will see why .

 

[trinibwoy]

But in the source engine its not as good of a buy as other cards.

We know why the Doom3 engine favors nvidia cards but why should source favor ati?

 

[Xmas]

Anyhoo...what's up with the jagged edges on the gun in the AA test for the 6800U?

I noticed that too; I was wondering why the reviewer didn't notice it. It looks pretty bad too.

The reviewer did notice.
However, I think that whole "IQ comparison" was pretty bad. A few slightly blurry upscaled cut-outs?

Anyway, the upper right edge looks better on ATI cards, the lower left edge looks better on NVidia cards. That's the effect of mirrored sample patterns.

 

[Randell]

does cz use heavy shading like HL2?

No way!

Condition Zero is based on the original Half-life engine.

True. But what CZ has to do with CS:S?

only people with CS:CZ are getting CS:S today.

 

[digitalwanderer]

only people with CS:CZ are getting CS:S today.

Not entirely correct, ATi HL2 voucher holders can also get it today.

I know, that's how I got it. 8)

 

[jvd]

But in the source engine its not as good of a buy as other cards.

We know why the Doom3 engine favors nvidia cards but why should source favor ati?

why did doom3 engine favor nvidia cards ?

We already know alot of things were done to make nvidia cards fast for doom3 engines . Both by nvidia and id.

So why can't we assume the same was done with the source engine .

we know valve spent alot of time optimizing for the fx line and it was still slower than the radeon line.

Do we really think the 6800 line closed the gap enough for there still not to be a lead by ati cards.

I don't think so.


But hey in a few weeks we will see the benchmarks

 

[Smurfie]

I was expecting a 30% increase from the X800XT over the 6800ultra(based on fillrate differences). At 1600x1200 with 4xaa and 8xaf, 70.54fps vs 62.1 is only a 14% increase, perhaps CPU bound? Good showing by the X800pro though, but it might actually show that this game is more fillrate than shader heavy.

Too many unknowns too, like optimizations in play and not. We need a proper review soon!

 

[BRiT]

The frame-rate cap is 90. I'm not sure if they benchmarked with it disabled. So that might be limiting the avg fps on some cards.

 

[jvd]

I was expecting a 30% increase from the X800XT over the 6800ultra(based on fillrate differences). At 1600x1200 with 4xaa and 8xaf, 70.54fps vs 62.1 is only a 14% increase, perhaps CPU bound? Good showing by the X800pro though, but it might actually show that this game is more fillrate than shader heavy.

Too many unknowns too, like optimizations in play and not. We need a proper review soon!

well first off the 30% was for half life 2. So you wont see it in this game. It is not as taxing as half life 2 will be.

The frame-rate cap is 90. I'm not sure if they benchmarked with it disabled. So that might be limiting the avg fps on some cards.

yea i was saying that too.

All the cards may actually do better with hte x800pro , xt , 6800gt , ultra doing alot better if they break 90fps more often than not

 

[termikk]

nvidia's counterparts to 420

 

[Randell]

only people with CS:CZ are getting CS:S today.

Not entirely correct, ATi HL2 voucher holders can also get it today.

I know, that's how I got it. 8)

true, damn annoying, it was at 93% about midnight, but I chose to go to bed, and now its done, I'm at work.