Inquirererer: ATi is teh Half-life2 benchy winnar!

[digitalwanderer]

Hanners posted up at EB about an article in the Inquirererer about ATi and a presentation of theres with ooodles of benchmarks showing how they wipe the floor with nVidia in Half-life2:

HALF LIFE 2 is finally with us and we've had sight of some benchmarks of ATI products with the long-awaited game.

We reproduce these below. The slides cover a number of different configurations used to play the game, and are pretty much self-explanatory.

I'm only posting this 'cause I took a gamebreak to move me Valve directory and defrag. (I spaced it and have been having bad load times. )

 

[Bjorn]

Though as with NVidias Doom3 benchmarks (x700 vs 6600GT f.e), let's perhaps wait until we have some independant benchmarks.

 

[RejZoR]

I mean does it really matters who wins? As long game is playable at reasonable settings...

 

[Mark]

I mean does it really matters who wins? As long game is playable at reasonable settings...

Where, exactly, do you think you are?

 

[Entropy]

Interesting, isn't it, that ATI only compares with the 6X00 series using 4xAA,8xAF, and for those graphs they don't give framerates, only percentages.

Hypothesis 1: nVidia performs relatively better at other settings than (4xAA,8xAF).
Hypothesis 2: The framerates achieved at the high resolution + (4xAA,8xAF) settings don't qualify as "playable", which is why they give percentages rather than fps. Using settings that give good game play, ATI does not have a performance advantage.

Further, note absence of specified driver version.

Frankly, if they need to resort to such crummy benchmarketing to demonstrate a performance advantage, it's reasonable to assume that nVidias products overall perform as well or better on Half-Life 2.

 

[Graphics_Krazy]

...or interesting in what slides the Inquirer posted

I don't know how you can spin this as a positive for NVIDIA.

 

[Entropy]

...or interesting in what slides the Inquirer posted

I don't know how you can spin this as a positive for NVIDIA.

I don't.
I spin it as a negative for ATI.

 

[Tim]

Interesting, isn't it, that ATI only compares with the 6X00 series using 4xAA,8xAF, and for those graphs they don't give framerates, only percentages.

I think you need some sleep, they have slides with FPS-scores and slides with percentage scores.

see:
http://www.theinquirer.net/images/articles/hal3.jpg

Edit:
I agree that it is a problem that the driver versions are missing thought.

 

[PatrickL]

Hehe entropy, you are losing credilbiity being so blatantly biased

 

[Entropy]

Hehe entropy, you are losing credilbiity being so blatantly biased

I admit I missed that graph.
Quite decent performance at those settings too!

 

[Tim]

Hehe entropy, you are losing credilbiity being so blatantly biased

I admit I missed that graph.
Quite decent performance at those settings too!

Yes all the high-end cards (from Ati or Nvidia) seems to have very playable performance at 1600x1200 with 4xAA and 8AF.

 

[dan2097]

Hehe entropy, you are losing credilbiity being so blatantly biased

I admit I missed that graph.
Quite decent performance at those settings too!

Yes all the high-end cards (from Ati or Nvidia) seems to have very playable performance at 1600x1200 with 4xAA and 8AF.

I dont know look at HL2-demo_canals_09 I wouldnt say 40fps AVERAGE is really enough, let along 30fps average, thats bound to be unconformtable as its bound to be in the 20s a lot of the time

 

[ellingsen1]

I am not suprised at the results but was a little suprised that ATI released the numbers with little info. I knew they were working closely with Valve and did not expect anything else. I heard they were even developing special levels for ATI users. I will wait for independent numbers before drawing any final conclusions. Who knows what driver version and settings they used. For all we know it could be the 61.77 driver that may not have many optimizations for the game.

 

[jb]

This is just a silly stunt done by ATI marketing. NV did the same thing with Doom3 numbers. Really if anyone puts any stock in this...then they need to have some sense beat into them with a club...or in this case a crowbar

 

[karlotta]

Interesting, isn't it, that ATI only compares with the 6X00 series using 4xAA,8xAF, and for those graphs they don't give framerates, only percentages.

Hypothesis 1: nVidia performs relatively better at other settings than (4xAA,8xAF).
Hypothesis 2: The framerates achieved at the high resolution + (4xAA,8xAF) settings don't qualify as "playable", which is why they give percentages rather than fps. Using settings that give good game play, ATI does not have a performance advantage.

Further, note absence of specified driver version.

Frankly, if they need to resort to such crummy benchmarketing to demonstrate a performance advantage, it's reasonable to assume that nVidias products overall perform as well or better on Half-Life 2.

the settings are the defualt for the game.

 

[Nick]

NVIDIA usually releases optimized drivers after a game is released. ATI usually releases bugfixes after a game is released.

Seriously now, I'll wait until the game is released and is benchmarked with retail cards and the newest drivers. I find it quite hard to believe that NVIDIA chips perform so well on Doom 3 but ATI performs up to 50% better on Half-Life 2. Or does this really have an explanation?

 

[Smurfie]

I mean does it really matters who wins? As long game is playable at reasonable settings...

Where, exactly, do you think you are?

My goodness, an admission?

Going to go off topic a moment.
Does anyone notice that when you alt-tab out of Half-life 2 and alt-tab back to the game, you get lots of white spots all over the screen? It actually stays and becomes extremely glaringly annoying while you move. I was observing this at my colleague's X800XTPE setup when he alt-tabbed out to try and run fraps. Anyone knows how this can be fixed? I believe his drivers are 4.9.

I do remember observing this phenomena too with my 9600 while playing City of Heroes and alt-tabbing. Doesn't happen all the time though.

 

[karlotta]

I mean does it really matters who wins? As long game is playable at reasonable settings...

Where, exactly, do you think you are?

My goodness, an admission?

Going to go off topic a moment.
Does anyone notice that when you alt-tab out of Half-life 2 and alt-tab back to the game, you get lots of white spots all over the screen? It actually stays and becomes extremely glaringly annoying while you move. I was observing this at my colleague's X800XTPE setup when he alt-tabbed out to try and run fraps. Anyone knows how this can be fixed? I believe his drivers are 4.9.

I do remember observing this phenomena too with my 9600 while playing City of Heroes and alt-tabbing. Doesn't happen all the time though.

No i dont. 4.12b drivers

 

[Fodder]

Does anyone notice that when you alt-tab out of Half-life 2 and alt-tab back to the game, you get lots of white spots all over the screen?

Haven't tried HL2, but if I alt-tab out of CS:Source all my shaders turn pink.

 

[Sabastian]

I've been playing HL2 on my system for the last few hours. Performance is not disapointing for me. Currently I'm using a XP2600 (Barton) 512 megs of system ram, Radeon 9600 xt, nForce 2 400 ultra with integrated soundstorm, HD 80 gig with an 8 meg buffer. I d/l the 4.12 beta (TC ATI Optimized Driver v1.7 Ultra) and the game play is great. I've no way to bench the game though. If there is a way I'd like to know if someone could tell me I'd post my frame rates just for shits and gigels.

The game is great BTW has much of the same charator as the original in terms of the feeling. The graphics are a notch above anything I've had the pleasure of playing before though. Definitly worth buying IMO. Really a great game.