HardOCP X1950XTX Crossfire review

[biostud]

http://enthusiast.hardocp.com/article.html?art=MTE0NCwxLCxoZW50aHVzaWFzdA==

 

[Ackmed]

Its early.

The gist of it is that in some cases the ATI Radeon X1950 XTX CrossFire platform, a dual-GPU solution, is outperforming 2x GeForce 7950 GX2?s in SLI, quad-GPU solution, in other cases they are even. That weighs in strongly for the performance that CrossFire can deliver.

Now, the first thought that comes to mind is that perhaps the NVIDIA Quad SLI drivers just aren?t mature enough yet in these games to really show us what it can do.

The second point is probably very true. I expect quad-SLI to get better as time goes on. All in all, competition is good.

The Bottom Line

ATI has proven they are a leader and not a follower with the X1950 XTX. ATI has released the world?s first consumer 3D graphics card with GDDR4 memory clocked at the highest ever stock speed that chews through games when it comes to high definition gaming. Memory bandwidth looks to one again be the defining factor in 3D performance. With a re-designed heatsink/fan unit, faster memory, and lowered price, the ATI Radeon X1950 XTX and CrossFire Edition are both serious 3D gaming video cards for the Hardcore that offer some value over NVIDIA?s more expensive 7950 GX2. ATI?s CrossFire dual GPU gaming platform looks to have just grown up.

 

[dug777]

Overall Gameplay Evaluation

Overall, with very few exceptions, the ATI Radeon X1950 XTX provided a noticeable and tangible improvement in the gameplay experience over the Radeon X1900 XTX. Compared to the BFGTech GeForce 7950 GX2 it even performed better than that in some games, while in others they were even. Simply put, the ATI Radeon X1950 XTX won more games and scenarios than it lost them compared to the BFGTech GeForce 7950 GX2.

Wow, never thought i'd see something so complimentary about ATI from that mob

Anyway, betcha they used teh default 'quality' setting into the bargain, turn on HQ, which imo is essential, and it would be looking far worse for nvidia...

 

[Ackmed]

Originally posted by: dug777


Anyway, betcha they used teh default 'quality' setting into the bargain, turn on HQ, which imo is essential, and it would be looking far worse for nvidia...

Driverheaven used them on one test, and had this to say about it;

The most noticeable aspect of the optimisations test is that the Radeon's see almost no performance reduction when enabling the highest quality settings in the driver, the Nvidia cards do show a significant performance hit and this allows the X1950XTX to close in to 8fps behind the 7950GX2 which as we all know is a much more expensive product.

http://www.driverheaven.net/reviews/X1950XTXreview/opts.htm

 

[BFG10K]

I really don't like this business of changing the settings for each GPU. After all that we still have no idea how much faster the X1950 is over the X1900.

I also love it how they think they are getting TrAA/AAA in Prey. :roll:

 

[Ackmed]

Other reviews;

http://www.bjorn3d.com/read.php?cID=954
http://www.driverheaven.net/reviews/X1950XTXreview/
http://www.hardwarezone.com/articles/control.php?cid=3&id=2020
http://www.hexus.net/content/item.php?item=6538
http://www.hothardware.com/viewarticle.aspx?articleid=861&cid=1
http://www.techpowerup.com/reviews/ATI/X1950XTX
http://www.theinquirer.net/default.aspx?article=33872

 

[tuteja1986]

that review is screwed :! WORTHLESS CRAP HARDCOP CRAP ... I WILL WAIT FOR A DECENT REVIEW

 

[BFG10K]

OMG, quieter than a 7900 GTX at idle and equal under load!

(7900 GTX SLI is still quieter under load than X1950 CF though).

 

[tuteja1986]

battlefield 2
Setting : 2048x1536x32 4xAA 16xAF
7950GX : 91.9FPS
X1950XT : 88.1FPS tie

Quake 4
Setting : 2048x1536x32 4xAA 16xAF
7950GX : 59.5 FPS big win
X1950XT : 37FPS

Call of duty 2
Setting : 2048x1536x32 4xAA 16xAF
7950GX : 34.2 FPS
X1950XT : 30 FPS
tie

Farcry HDR
Setting : 2048x1536x32 0xAA 0xAF
7950GX : 60FPS big win
X1950XT : 43.3FPS

FEAR
Setting : 2048x1536x32 4xAA 16xAF
7950GX : 46FPS big win
X1950XT : 37FPS

HL2 Loast Coast HDR
Setting : 2048x1536x32 4xAA 16xAF
7950GX : 55 FPS big win
X1950XT : 41.8 FPS

Oblivion Mountain
Setting : 2048x1536x32 4xAA 16xAF
7950GX : 39.6 FPS
X1950XT : 34.8 FPS tie

Oblivion Foliage Area
Setting : 2048x1536x32 4xAA 16xAF
7950GX : 22.1 FPS
X1950XT : 21.8 FPS TIE

Overclocking performance increase 7% average

Conculsion ATI DOESN"T FREAKEN defeat the 7950GX in QUality Testing but i .don't know about High Quality has i haven't seen reliable review yet.

BF2 , COD 2 and Oblivion ATI does really well with mini frame rate but then again it just ties with 7950GX. I recommend you guys getting a X1950XTX if you can find it for $400 otherwise go with a $500 7950GX.

 

[BFG10K]

.don't know about High Quality has i haven't seen reliable review yet.

Follow my link - they do HQ on nVidia in all of the tests.

 

[tuteja1986]

Originally posted by: BFG10K

.don't know about High Quality has i haven't seen reliable review yet.

Follow my link - they do HQ on nVidia in all of the tests.

I just skimmed through it , i couldn't read the text ;( but i think it would have been even more awesome review if they did a higher res than 1600x1200 and also bechmarked with 8xAA/16xAA or even 14xATI's Adaptive AA vs Nvidia SLI 12x AA.

 

[tuteja1986]

I wish one website would just bechmark higher res than 1600x1200 and forget about the low res and just spend more time on 8x AA ++ than 4xAA. Most website have the same crappy review ;(

 

[Ackmed]

Originally posted by: tuteja1986
that review is screwed :! WORTHLESS CRAP HARDCOP CRAP ... I WILL WAIT FOR A DECENT REVIEW

They do things most do not. Such as, use HDR+AA, HQ AF, TRAA and AAA, 1920x1200/2560x1600, and actually talk about IQ, with pictures to compare. Call it worthless crap if you want, I like some of their settings.

 

[tuteja1986]

hey BFG :! what the hell is up with hexus review ;(

http://www.hexus.net/content/item.php?item=6538&page=5

when did they become ATI FANIDIOT WEBSITE :? did they make up the result or ?

 

[tuteja1986]

Originally posted by: Ackmed

Originally posted by: tuteja1986
that review is screwed :! WORTHLESS CRAP HARDCOP CRAP ... I WILL WAIT FOR A DECENT REVIEW

They do things most do not. Such as, use HDR+AA, HQ AF, TRAA and AAA, 1920x1200/2560x1600, and actually talk about IQ, with pictures to compare. Call it worthless crap if you want, I like some of their settings.

but they compare it with different setting noob don't look at the setting really :! they look at score and they tend to favour Nvidia with setting they use.

 

[LittleNemoNES]

lol HardOCPs playable setting are funny. Those cards, expecially in crossfire/SLI are capable of so much more. They use 6xAAA I use 14xAAA @ 1680x1050...lol. I guess they like getting 500fps. HardOCPs benching is flawed. I know that for me, lowering the Actor fade distance in oblivion to 75% gives a 10 fps increase without affecting gameplay. Most actors are indoors anyway but even then its not like they appear 10ft in front of you.

 

[BFG10K]

but i think it would have been even more awesome review if they did a higher res than 1600x1200

The higher res results are at the bottom of each page.

when did they become ATI FANIDIOT WEBSITE :? did they make up the result or ?

I'm not sure what you mean. If you mean the dark brown bar way ahead of the others keep in mind it's crossfire.

 

[TroubleM]

Seems to me that 7950GX2 has some serious problems on Conroe setups. They don't work as good compared to the AMD platforms even using nforce chipsets. That's how it can be explained the first benchmarks that put x1950xtx in front.

 

[tuteja1986]

Originally posted by: BFG10K

but i think it would have been even more awesome review if they did a higher res than 1600x1200

The higher res results are at the bottom of each page.

when did they become ATI FANIDIOT WEBSITE :? did they make up the result or ?

I'm not sure what you mean. If you mean the dark brown bar way ahead of the others keep in mind it's crossfire.

ohh i didn't notice that they should really add more info

 

[Ackmed]

Originally posted by: tuteja1986

Originally posted by: Ackmed

Originally posted by: tuteja1986
that review is screwed :! WORTHLESS CRAP HARDCOP CRAP ... I WILL WAIT FOR A DECENT REVIEW

They do things most do not. Such as, use HDR+AA, HQ AF, TRAA and AAA, 1920x1200/2560x1600, and actually talk about IQ, with pictures to compare. Call it worthless crap if you want, I like some of their settings.

but they compare it with different setting noob don't look at the setting really :! they look at score and they tend to favour Nvidia with setting they use.

Most of the time yes, they are different settings. Whos fault is it if they dont read, and just look at the pictures?

Going by most reviews out there, people wouldnt have a clue that NV has 8xAA, or that ATi has HQ AF, can do HDR+AA in games when NV cant, or that both have better AA to clean up the image with TRAA and AAA, etc. I dont think that HardOCPs testing is perfect, but they do use settings that I use. I use HDR+AA, I use HQ AF, I use AAA. Who else tests with these settings? I do wish they would put back the "apples to apples" comparisons, but thats not really true either, because like most, they dont use HQ for NV, and leave ops on. Dismissing them totally is pretty shortsighted however, imo.

Originally posted by: gersson
lol HardOCPs playable setting are funny. Those cards, expecially in crossfire/SLI are capable of so much more. They use 6xAAA I use 14xAAA @ 1680x1050...lol. I guess they like getting 500fps. HardOCPs benching is flawed. I know that for me, lowering the Actor fade distance in oblivion to 75% gives a 10 fps increase without affecting gameplay. Most actors are indoors anyway but even then its not like they appear 10ft in front of you.

Its a matter of opinion, and thats where some of the fault lies. They do use 14xAA in games, that they feel they can get playable frames. You dont get playable frames at 14xAA all the time either. Getting 500fps? Not hardly, try not to exaggerate to help your arguement.

 

[Ackmed]

ATs review is up; http://www.anandtech.com/video/showdoc.aspx?i=2821

FS review; http://www.anandtech.com/video/showdoc.aspx?i=2821

OP.. change title to include all reviews, to make it easier.

 

[josh6079]

Wow, doesn't look like much of a performance benefit from an X1900XTX. Only when you use crossfire does it look like it bumps things up.

 

[nitromullet]

Looks like a nice card. The reviews are kinda all over the place though. I tend to believe FS and AT the most out of what I've seen, showing the X1950XTX as the overall fastest single gpu card, but not a huge increase over the X1900XTX.

 

[JackBurton]

What I was concentrating on was the 2560x1600 scores. And they look good. Nice job ATi!

 

[zephyrprime]

Take a look at some of the individual feature tests from the inquirer's test:

Composite Figures 3Dmark 05 Single Texturinng Multi Textur. Pixel Shader
VS/VS
complex
ATI X1900XTX..........650/1550 MHz...5060.8 10237.4 439.4 141.8/61.9
ATI X1950XTX..........650/2000 MHz...6307.2 10209.2 456.6 147.9/61.9
ATI X1950XTX OC...700/2200 MHz...6711.2 10971.8 475.1 154.5/62.3

As you can see, the single texturing performance is scaling roughly linearly with memory clock speed. Multitexturing showed no improvement at all which I think is surprising. Pixel shading is little changed which is to be expected.

Anyone have any ideas on why multitexuring performance would see no improvement while single texturing saw a big improvement?