Half Life 2 Benchmarks @ HardOCP.com

[Brent_Justice]

Great review. I'm enjoying HL2 at 1600x1200 with frame rates always near 100.
Anyone with an older card, upgrade, make that your early Christmas present or Black Friday gift for next week.
D3 was indoors and dark, so you couldn't see much. But HL2 is outside, and the views of the city are incredible.

I have one question. In the advanced display settings of the video card, do you have the slider set to High Performance or Quality? I assume Performance, because it can make a differance of 10 FPS. I have both boxes checked on application controlled.

"Quality" which is the default setting of the driver.

 

[fallguy]

I'm actually quite surprised ATI doesn't have a significant performance lead while operating at 100mhz faster clock speed and at lower shader precision with less features on chip.

Posts like this is why CM doesnt post here very often. Please dont try to troll.

 

[tranCendenZ]

Posts like this is why CM doesnt post here very often. Please dont try to troll.

If his post wasn't pure PR maybe I wouldn't have responded in such a manner And its not trolling to challenge PR.

 

[Brent_Justice]

Out of the box experience? LOL!
It's retarded for any self-respecting benchmarker to publish a graphics card benchmark which is running CPU-limited tests.
Why haven't you written benchmarks based on "out of box experience" for Far Cry, DOOM3 etc ?

Our demos reflect real world gameplay in two maps, througout the ENTIRE level. They give a good indication of what a gamer experiences actually playing HL2. They involve many things, for example Canals has smoke, exploding barrels, gun fire, and water. Coast has physics with the buggy, water, smoke, gun fire, the splashes from the water, lots of explosions with the air ship at the end. It's everything a gamer experiences while playing the game.

We did a DOOM 3 Hardware Guide. We wanted to do a HL2 Hardware Guide, but unfortuntely could not do one before HL2 was released.

 

[Brent_Justice]

If his post wasn't pure PR maybe I wouldn't have responded in such a manner And its not trolling to respond to PR.

Play nice now.

 

[CATALYST MAKER]

If his post wasn't pure PR maybe I wouldn't have responded in such a manner And its not trolling to challenge PR.

How is my post pure PR? You dont mean to tell me that you dont clearly realize from all the reviews across the web that HL2 is quite a CPU bound game. In scenes where that is true, ATI and NV are tied. However in scenes that this is not the case, ATI is a significant winner in FPS?

I dont know about you but I prefer all the eye candy cranked.

The following was not written by me.....

"Update: We have gone back and updated our pages with a couple of graphs that show Maximum IQ settings in terms of AA and AF. Without a doubt the ATI Radeon X800XT-PE did by the far the best job at delivering a playable gaming experience. Of course it is up to the end user to determine if turning these options on give you any tangible gaming returns, but without a doubt if you want to run "ultra high quality settings," the ATI Radeon X800XT-PE gives a much better return than NVIDIA's solution."

 

[pahncrd]

How is my post pure PR? You dont mean to tell me that you dont clearly realize from all the reviews across the web that HL2 is quite a CPU bound game. In scenes where that is true, ATI and NV are tied. However in scenes that this is not the case, ATI is a significant winner in FPS?

I dont know about you but I prefer all the eye candy cranked.

The following was not written by me.....

"Update: We have gone back and updated our pages with a couple of graphs that show Maximum IQ settings in terms of AA and AF. Without a doubt the ATI Radeon X800XT-PE did by the far the best job at delivering a playable gaming experience. Of course it is up to the end user to determine if turning these options on give you any tangible gaming returns, but without a doubt if you want to run "ultra high quality settings," the ATI Radeon X800XT-PE gives a much better return than NVIDIA's solution."

I probably would have an ATi card if they had the features I want. Speed isn't everything.

Right now I am playing HL2 with 8xSSAA at 1280x960, and it barely dips below 40fps in the most extreme cases. It's ussually above 60fps.
Not to mention it is shaping up to be pretty amazing in stereo 3d. Hopefully nvidia releases the new stereo drivers soon.

So as soon as you can play on a radeon with 3d stereo or at 8xSSAA, I will be jealous and convert.

 

[karnick]

Kinda sad considering the x800 needs to have its clocks speeds ramped way up just to stay a couple frames ahead of the NV40...now clock that NV40 as high as the x800 or down clock the x800 to the speed of an ultra and see who gets whos ass handed to em.

go clock an AMD chip at the same clock speed as an intel........ clock speed doesnt have everything to do with it

 

[CastleBravo]

*Had* a performance lead?

Ohh you must be talking about being tied in CPU bound cases....

CPU bound with an FX-53 test rig running 1600x1200 4xAA 8xAF?

You realize that the x800 Pro remains inferior to the 6800 GT even here, the one place it was FINALLY supposed to show us something, in the DX9 game it was supposed to "own," with lots of AA and AF that it is supposed to do faster. Oops. Still a loser. But at least ATI fans can console themselves that they can beat a widely-available $500ish 6800 Ultra with a borderline-vaporware x800 XT-PE going for $100+ more... excepting a free download of Coolbits 2.0 by the Ultra owner, of course.

 

[coz]

Terry, could you please explain (for the benefit of myself and the others here) why the option "+r_fastzreject 1" isn't enabled by default on the X800 cards but it is on the 6800 cards? Maybe that's more a question for Valve but I'm sure you know the answer anyway o' great Maker!

 

[CATALYST MAKER]

CPU bound with an FX-53 test rig running 1600x1200 4xAA 8xAF?

You realize that the x800 Pro remains inferior to the 6800 GT even here, the one place it was FINALLY supposed to show us something, in the DX9 game it was supposed to "own," with lots of AA and AF that it is supposed to do faster. Oops. Still a loser. But at least ATI fans can console themselves that they can beat a widely-available $500ish 6800 Ultra with a borderline-vaporware x800 XT-PE going for $100+ more... excepting a free download of Coolbits 2.0 by the Ultra owner, of course.

Correct.

But what are you talking about? The X800 Pro is inferior to the 6800 GT?

Let me guess you own one right?

http://www.firingsquad.com/hardware/half-life_2_performance/page8.asp

 

[Moloch]

CPU bound with an FX-53 test rig running 1600x1200 4xAA 8xAF?

You realize that the x800 Pro remains inferior to the 6800 GT even here, the one place it was FINALLY supposed to show us something, in the DX9 game it was supposed to "own," with lots of AA and AF that it is supposed to do faster. Oops. Still a loser. But at least ATI fans can console themselves that they can beat a widely-available $500ish 6800 Ultra with a borderline-vaporware x800 XT-PE going for $100+ more... excepting a free download of Coolbits 2.0 by the Ultra owner, of course.

HL2= major physics, the X800 pro is about the same speed as the GT when run at ref. clock speed, while offering better FSAA options like temporal AA, which doto the high framerates in HL2, is a possability, where 6X and 8X on nvidia hardware is not, unless you like playing at 30-40fps.

 

[CastleBravo]

Correct.

But what are you talking about? The X800 Pro is inferior to the 6800 GT?

Let me guess you own one right?

http://www.firingsquad.com/hardware/half-life_2_performance/page8.asp

Gee, let me think. If a video card loses practically every benchmark under the sun on practically every review site to a another card, can I call the former "inferior" to the latter or not? Oh, and don't forget that they cost the same and are both widely available, unlike the situation with the x800 XT-PE vs. the 6800 Ultra... oops.

A small sample of x800 Pro loses to the 6800 GT…

http://www.hardocp.com/article.html?art=NjkyLDE=
http://graphics.tomshardware.com/graphic/20041004/vga_charts-04.html
http://graphics.tomshardware.com/graphic/20041004/vga_charts-07.html
http://graphics.tomshardware.com/graphic/20041004/vga_charts-08.html
http://www.hardocp.com/article.html?art=NjQ0
http://www.anandtech.com/video/showdoc.aspx?i=2278&p=11
http://www.anandtech.com/video/showdoc.aspx?i=2146&p=3
http://www.xbitlabs.com/articles/video/display/cs-source_14.html
http://www.xbitlabs.com/articles/video/display/counterstrike-source_11.html
http://www.xbitlabs.com/articles/video/display/doom-3-tests_8.html
http://www.xbitlabs.com/articles/video/display/graphics-cards-2004_17.html
http://www.xbitlabs.com/articles/video/display/graphics-cards-2004_21.html
http://www.xbitlabs.com/articles/video/display/graphics-cards-2004_28.html
http://www.xbitlabs.com/articles/video/display/farcry13_19.html Look, the x800 Pro wins! Oh, wait, that’s 1024x768, the GT beat it at all the higher resolutions tested…

 

[pahncrd]

HL2= major physics, the X800 pro is about the same speed as the GT when run at ref. clock speed, while offering better FSAA options like temporal AA, which doto the high framerates in HL2, is a possability, where 6X and 8X on nvidia hardware is not, unless you like playing at 30-40fps.

8X on nvidia at 60+ fps is very possible for me at 1280x960.

 

[Moloch]

8X on nvidia at 60+ fps is very possible for me at 1280x960.

On maps that aren't really that hard on the GPU, try doing that on the canals map.
http://www.firingsquad.com/hardware/half-life_2_performance/page8.asp
But to be fair, you cant use TAA either, but you can use 6X FSAA, whick looks loads better than 4x.
If HL2 isn't GPU bound, how come there isn't a large difference between the X800 pro and the XT, like there in other DX9 level games?
The XT-PE has a fillrate of 83XX vs the X800 pro's 5700, not to meation a memory lock of 1120 vs 900 for the pro, and of course 16 pipelines- more shader power .

 

[priyajeet]

If his post wasn't pure PR maybe I wouldn't have responded in such a manner And its not trolling to challenge PR.

Even if it pure PR, its OK. I dont see his NV counterparts helping/interacting with the community. Something is better than nothing.

When D3 came out, people like HUMUS helped out the ati community. Wud like to see NV guys do the same. I think ATI has done a pretty good job of interacting with the users.

 

[Jonsey]

Even if it pure PR, its OK. I dont see his NV counterparts helping/interacting with the community. Something is better than nothing.

What do you mean we have no nvidia PR? We have tranCendenZ after all!

 

[pahncrd]

On maps that aren't really that hard on the GPU, try doing that on the canals map.
http://www.firingsquad.com/hardware/half-life_2_performance/page8.asp
But to be fair, you cant use TAA either, but you can use 6X FSAA, whick looks loads better than 4x.
If HL2 isn't GPU bound, how come there isn't a large difference between the X800 pro and the XT, like there in other DX9 level games?
The XT-PE has a fillrate of 83XX vs the X800 pro's 5700, not to meation a memory lock of 1120 vs 900 for the pro, and of course 16 pipelines- more shader power .

I did play on canals with the fps staying around 40-50fps. I am using the 67.02 drivers with opts on.

I have a A64 939 at 250x10 2.5ghz, and 6800 gt at 430/1100

I really don't see why anyone is complaining.

 

[Moloch]

Even if it pure PR, its OK. I dont see his NV counterparts helping/interacting with the community. Something is better than nothing.

It's not bad PR if it's true- check out the canals map, the X800 pro is beating the ultra.
But that's an old driver firingsquad is using, and nvidia "fixed" performance.
I'd wait till the next driver version for both companies are available before saying who is better, even without the fastzreject on.

 

[Moloch]

I did play on canals with the fps staying around 40-50fps. I am using the 67.02 drivers with opts on.

I have a A64 939 at 250x10 2.5ghz, and 6800 gt at 430/1100

I really don't see why anyone is complaining.

40-50 is kind of low for a FPS, most start to feel laggy around 40-50, and if it's 40-50fps, there must be some drops to the 30s or lower during heavy battle, btw how many fps do you get at default clockrate, apples to apples please.

 

[pahncrd]

40-50 is kind of low for a FPS, most start to feel laggy around 40-50, and if it's 40-50fps, there must be some drops to the 30s or lower during heavy battle, btw how many fps do you get at default clockrate, apples to apples please.

My point was that i am enjoying features unavailable on Nvidia only. I would probably switch if that were not the case.
I certainly don't feel any mouse lag, and it never dips below 35fps or so in extreme firefights. Though there is an occasional 'chug' as there is a lot of physics.
I will go through some action parts and take a few screens.

 

[CrimandEvil]

Don't pay attention to Moloch, you'll learn to disregard his posts as mostly FUD or flamebaiting.

 

[^eMpTy^]

Correct.

But what are you talking about? The X800 Pro is inferior to the 6800 GT?

Let me guess you own one right?

http://www.firingsquad.com/hardware/half-life_2_performance/page8.asp

Yeah, inferior...as in, it's slower...surely you've noticed by now that not one single review of the x800pro has recommended it over the 6800GT...well...I guess maybe DH might have...

Nice pick of benchmarks...why don't you just link to the same article you quoted before...hmm...maybe because in the [H] benchmarks the Pro is slower than the GT? So the [H] is only right when they support your cause?

 

[Moloch]

My point was that i am enjoying features available on Nvidia only. I would probably switch if that were not the case.
I certainly don't feel any mouse lag, and it never dips below 35fps or so in extreme firefights. Though there is an occasional 'chug' as there is a lot of physics.
I will go through some action parts and take a few screens.

Fixed, the chugs are what I'm talking about, the min framerate is more important than the avg framerate, unless it's doing it during the first few secs of the map(still loading).

Don't pay attention to Moloch, you'll learn to disregard his posts as mostly FUD or flamebaiting.

Yup, telling it how it is is FUD and flaming.. right on

Yeah, inferior...as in, it's slower...surely you've noticed by now that not one single review of the x800pro has recommended it over the 6800GT...well...I guess maybe DH might have...

Nice pick of benchmarks...why don't you just link to the same article you quoted before...hmm...maybe because in the [H] benchmarks the Pro is slower than the GT? So the [H] is only right when they support your cause?

Brent decided to use BFG cards that aren'te at nvidia recommended/ referance clock speeds. For anyone who doesn't buy that specific brand, or gainward cards, they will have slightly lower fps.
All others show the pro either slightly ahead, or tied.

 

[^eMpTy^]

Fixed, the chugs are what I'm talking about, the min framerate is more important than the avg framerate, unless it's doing it during the first few secs of the map(still loading).

Yup, telling it how it is is FUD and flaming.. right on

Brent decided to use BFG cards that aren'te at nvidia recommended/ referance clock speeds. For anyone who doesn't buy that specific brand, or gainward cards, they will have slightly lower fps.
All others show the pro either slightly ahead, or tied.

lol...this has been explained a hundred times already...the BFG cards are the most popular ones...more people have them than any other card...

strange how you don't seem to be upset that they used an XTPE vs a vanilla XT...the XTPE isn't even available in retail for under $700...why aren't you angry about that Moloch?

and who cares about HL2...it's just one game right? I'd much rather have a GT which is faster in almost every game...STUPID faster in D3...and tied in HL2...call me crazy...

 

[pahncrd]

Fixed, the chugs are what I'm talking about, the min framerate is more important than the avg framerate, unless it's doing it during the first few secs of the map(still loading).

The chugs are from a lot of physics at once. It happens regardless of settings.

I just ran through water hazard and at no time did my fps dip below 30, but it did hit 30 once. My average was 56.6 fps, and this is with AF at 4x and clocks set at 400/1100.

 

[^eMpTy^]

The chugs are from a lot of physics at once. It happens regardless of settings.

the chugs are from auto-saves...ever notice that if you load from an auto-save it was always at a point you remember the framerate dropped real bad?

 

[pahncrd]

the chugs are from auto-saves...ever notice that if you load from an auto-save it was always at a point you remember the framerate dropped real bad?

Hmm, good point, I always assumed something was just blowing up .

 

[Jonsey]

strange how you don't seem to be upset that they used an XTPE vs a vanilla XT...the XTPE isn't even available in retail for under $700...why aren't you angry about that Moloch?

and who cares about HL2...it's just one game right? I'd much rather have a GT which is faster in almost every game...STUPID faster in D3...and tied in HL2...call me crazy...

http://www.wallapc.com/.sc/ms/dd/11...X800 XT PLATINUM EDITION 256MB AGP WHITE BOX

I don't know if the link works, but it's a XT PE in stock for 545. (I don't know if they'll last long, though.)

Oh, by the way, you're crazy, ^eMpTy^

 

[fallguy]

Gee, let me think. If a video card loses practically every benchmark under the sun on practically every review site to a another card, can I call the former "inferior" to the latter or not? Oh, and don't forget that they cost the same and are both widely available, unlike the situation with the x800 XT-PE vs. the 6800 Ultra... oops.

A small sample of x800 Pro loses to the 6800 GT…

You realize that the x800 Pro remains inferior to the 6800 GT even here, the one place it was FINALLY supposed to show us something, in the DX9 game it was supposed to "own," with lots of AA and AF that it is supposed to do faster. Oops. Still a loser.

The X800XT is faster than the Ultra in the res+settings I use. Its cheaper too.

Yes the XT/PE is hard to get, but people do have them, and you can get them. Going by your logic of overclocking with coolbits for the Ultra, you can do the same with the XT.

"Still a loser"? Any non bias person who looked at the reviews wouldnt call the X800 PRO a loser compared to the GT.

pahncrd, I tried 8x as well, and it was very slow for me. Our CPU's and GPU's are the same, same drivers as well. I wasnt anywhere near 40-50 for an average. I was at 1280x1024 though, which is pretty much the same as the res you used.

 

[pahncrd]

The X800XT is much faster than the Ultra in the res+settings I use. Its cheaper too.

Yes the XT/PE is hard to get, but people do have them, and you can get them. Going by your logic of overclocking with coolbits for the Ultra, you can do the same with the XT.

"Still a loser"? Any non bias person who looked at the reviews wouldnt call the X800 PRO a loser compared to the GT.

pahncrd, I tried 8x as well, and it was very slow for me. Our CPU's and GPU's are the same, same drivers as well. I wasnt anywhere near 40-50 for an average. I was at 1280x1024 though, which is pretty much the same as the res you used.

I can still post screens, though it is a bit of a pain in the ass. I have no reason to mislead. That is simply what I play on, and what I get in fps. I play with opts on.

Also, what are your cpu/gpu clocks?

I am sure everyone is missing my point, in that the features that I want are on Nvidia, and 10 more fps in a game at 16x12 just doesn't do it for me.
If ATi had 3d stereo and SSAA I would own their card. In this game in particular, I AM able to run at simply amazing image quality.

 

[Moloch]

lol...this has been explained a hundred times already...the BFG cards are the most popular ones...more people have them than any other card...

strange how you don't seem to be upset that they used an XTPE vs a vanilla XT...the XTPE isn't even available in retail for under $700...why aren't you angry about that Moloch?

and who cares about HL2...it's just one game right? I'd much rather have a GT which is faster in almost every game...STUPID faster in D3...and tied in HL2...call me crazy...

Start a poll in the nvidia sub forum asking which card maker people have.
I can't believe you think that 2 card makers that I know of that have overclocked stocked gpus is better than using reference clockspeeds, every X800 pro or XT is gonna be running at the same clockspeed, whereas only a few nvidia makers offer overclocked stock cards for a price premium.
700 dollers? A few places are still selling em for that, but I can see some for under 500 here http://www.pricegrabber.com/search_getprod.php/masterid=4751901/search=X800 xt
http://www.pricegrabber.com/search_getprod.php/masterid=4927909/search=X800 xt
And guess what? Both run at the same clockspeed, and before you say that's only 2 cards before clicking link, its 2 different cards and listing where you can get them at.

 

[^eMpTy^]

http://www.wallapc.com/.sc/ms/dd/11...X800 XT PLATINUM EDITION 256MB AGP WHITE BOX

I don't know if the link works, but it's a XT PE in stock for 545. (I don't know if they'll last long, though.)

Oh, by the way, you're crazy, ^eMpTy^

From their website:

RETAIL BOX MANUFACTURE WARRANTY ONLY. NO RETURN OR EXCHANGE.
ONLY FEW IN-STOCK. IF BACK ORDERED PLEASE DO NOT ORDER.

Also from their website:

SAPPHITE ATI X800 XT PLATINUM EDITION 256MB AGP WHITE BOX

So...is it a white box? or is it retail? is it in stock? or is it backordered?

I tell you what, if you have the balls to order from that place...more power to you, until the cards show up on newegg...then they aren't in retail in my book...

http://www.newegg.com/app/viewProductDesc.asp?description=14-102-438&depa=0

 

[Jonsey]

From their website:

RETAIL BOX MANUFACTURE WARRANTY ONLY. NO RETURN OR EXCHANGE.
ONLY FEW IN-STOCK. IF BACK ORDERED PLEASE DO NOT ORDER.

Also from their website:

SAPPHITE ATI X800 XT PLATINUM EDITION 256MB AGP WHITE BOX

So...is it a white box? or is it retail? is it in stock? or is it backordered?

I tell you what, if you have the balls to order from that place...more power to you, until the cards show up on newegg...then they aren't in retail in my book...

http://www.newegg.com/app/viewProductDesc.asp?description=14-102-438&depa=0

Well that was just the first one I saw on price watch. I didn't realize it had to be at newegg to meet your definition of retail though. I'll keep that in mind in the future.

 

[^eMpTy^]

Start a poll in the nvidia sub forum asking which card maker people have.
I can't believe you think that 2 card makers that I know of that have overclocked stocked gpus is better than using reference clockspeeds, every X800 pro or XT is gonna be running at the same clockspeed, whereas only a few nvidia makers offer overclocked stock cards for a price premium.
700 dollers? A few places are still selling em for that, but I can see some for under 500 here http://www.pricegrabber.com/search_getprod.php/masterid=4751901/search=X800 xt
http://www.pricegrabber.com/search_getprod.php/masterid=4927909/search=X800 xt
And guess what? Both run at the same clockspeed.

Wow...way to miss the point...I said PLATINUM EDITION...

You start a poll, you're the one freaking out about it...everyone else was done talking about it in this thread like 15 pages ago...

I never said that anything was better than anything else...where do you get this nonsense?

All I said, was what Brent said, that the OC versions are very popular retail cards and therefore should be in the test...given the price and availability of the XT PLATINUM EDITION (can you read it this time?) you're lucky they didn't put it up against a waterblocked Ultra running at 450Mhz...

 

[^eMpTy^]

Well that was just the first one I saw on price watch. I didn't realize it had to be at newegg to meet your definition of retail though. I'll keep that in mind in the future.

So go find one...don't give up so easily...prove me wrong...find one etail store who's name I've heard of (and I've heard of a lot) with the card in stock for under $700 (no, 699 doesn't count)...

I mean seriously...this is common knowledge at this point...why dispute it? the XTPE just got replaced with the vanilla XT because ATi only ever made the XTPE for press purposes...why do you think all their vaunted OEM sales are of vanilla XTs???

 

[CrimandEvil]

LOL, it doesn't have to be Newegg, as long as there not some "fly by night" retailer.

 

[Moloch]

Wow...way to miss the point...I said PLATINUM EDITION...

You start a poll, you're the one freaking out about it...everyone else was done talking about it in this thread like 15 pages ago...

I never said that anything was better than anything else...where do you get this nonsense?

All I said, was what Brent said, that the OC versions are very popular retail cards and therefore should be in the test...given the price and availability of the XT PLATINUM EDITION (can you read it this time?) you're lucky they didn't put it up against a waterblocked Ultra running at 450Mhz...

The XT-PE is akin to the 6800 ultra extreme..
The XT will still win benchmarks vs the 6800 ultra.

 

[Jonsey]

So go find one...don't give up so easily...prove me wrong...find one etail store who's name I've heard of (and I've heard of a lot) with the card in stock for under $700 (no, 699 doesn't count)...

I mean seriously...this is common knowledge at this point...why dispute it? the XTPE just got replaced with the vanilla XT because ATi only ever made the XTPE for press purposes...why do you think all their vaunted OEM sales are of vanilla XTs???

Want to give me a list of the stores you have heard of? So does $698 count? How about $697? $696?

I'm not really serious of course. I just can't help taking a thread off course when someone makes a dogmatic statement. I realize the X800 XT PE is very hard to find at reasonable prices, I won't argue that with you. You don't need to give me the dreaded "roll eyes."

 

[^eMpTy^]

The XT-PE is akin to the 6800 ultra extreme..
The XT will still win benchmarks vs the 6800 ultra.

In HL2 only...and by the same token, the performance difference between the Ultra and Ultra OC is small as well...so the question remains...why don't you just let it go?