[H]ardocp Review for the Geforce FX

[Tahir2]

The jury is still reviewing the evidence but I was expecting better AA IQ if nothing else.

 

[nggalai]

Does Nvidia use 4X rotated grid MS?

Nope.

ta,
-Sascha.rb

 

[mr]

Is it just me or is the GFFX missing fog not only in NFS but in UT2003 too?
If you look at the AF comparison in [H]'s review the Anubis map has no fog on both cards so I presume there is none, but both the suntemple map and the HQ shots show fog on the R300 and none on GFFX.

I don't know UT2003 "in depth" to judge which one looks right, but they sure look different regarding fog. It is also difficult to see if fog is missing in these shots, or simply further in the distance.

I can't find proper screenshots of any games besides those at [H], so I guess we will have to wait for the beyond3d review.

I'm sorry if this is already covered somewhere, but I'm unable to find any comment regarding fog in UT2003.

 

[Nappe1]

Is it just me or is the GFFX missing fog not only in NFS but in UT2003 too?
If you look at the AF comparison in [H]'s review the Anubis map has no fog on both cards so I presume there is none, but both the suntemple map and the HQ shots show fog on the R300 and none on GFFX.

I don't know UT2003 "in depth" to judge which one looks right, but they sure look different regarding fog. It is also difficult to see if fog is missing in these shots, or simply further in the distance.

I can't find proper screenshots of any games besides those at [H], so I guess we will have to wait for the beyond3d review.

I'm sorry if this is already covered somewhere, but I'm unable to find any comment regarding fog in UT2003.

it looks like to be "Table Fog" problem... 3dfx was the last one doing it fully HW. (afaik. correct me if I am wrong. ATI, Matrox as well as nVidia have had it as software emulation.)

How about Radeon 9700?? does it use Pixel Shaders for doing Table fog on DX7/DX6 apps?? I was suprised how well and without penalty hit it does it even on NFS4. (Since Voodoo2 I haven't seen so fluid Table Fog. All the others so far have had software emulation for it and it have been working more or less well...)

 

[Ailuros]

One oddity.

The villagemark scores show the fillrate advantage nicely on Toms site. It also refutes the most common belief that the bench is a test of bandwidth saving features. Not true, it tests fillrate and the ability of the card to discard pixels.

Its also one of the first time in recent memory that an Nvidia card has performed well on that bench.

A bit OT: the GF4 here jumps from ~120fps in 1024x768x32 T&Lon to ~145fps with T&Loff.

It's worth a shot even for Radeon users

 

[MuFu]

Another great review from Brent.

Originally posted by Brent
An option that may be somewhat confusing is the "Texture Sharpening" check box. If you enable this checkbox it will set the Anisotropic Level up 1 notch. For example if you are at 2X Anisotropic Filtering and you enable Texture Sharpening it will now operate at 4X AF. If you are at 4X AF and enable Texture Sharpening it will run at 8X AF. If you are already at 8X AF then turning on Texture Sharpening will do nothing. It doesn’t have anything to do with LOD. It works if you have AA enabled or disabled. The theory behind this option was explained to us a being there for those folks that have no idea what Anisotropic Filtering is and are afraid to increase the AF slider. This way even scared n00bs get some Aniso love as everyone most likely knows that things look "better" with sharper textures.

*round of applause*

Originally posted by FrgMstr
Brent went to bed....

You are speaking of a real possibility here considering they are right next to each other. Good point.

I was getting exhaust temps (I had the card here running for a few hours on Friday) at around 120F running the card in DOS.

Hmm... that's 50C when idling! WOWSER.

Amazing to think that ATi could release a "dustbuster-equipped" 9700 Pro right now that would take the performance crown back immediately, possibly by a considerable margin. That said, I really hope they do not use the 5800 Ultra as an excuse to launch R350 with a similar contraption. People have been saying "they have to in order to get to 400MHz!" but I really don't think that's true.

Also interesting that the fastest card available right now is probably not the GeForceFX 5800 Ultra, but one of the AIB 9700 Pro boards. Hell... there's one with passive cooling that still costs less than FX.

MuFu.

 

[demalion]

Hmm...the content of that review strikes me as excellent, but some of the structure needs work, perhaps because they were rushed?

I still am not sure if they specifically mentioned what the aniso settings were in terms of aggressive/balanced etc in detail (though it is clear what they are using by looking at the final comparison image).

Some creativity in the screen shot presentation would improve the review quite a bit...perhaps associating screenshots with graph results in some way.

I really like the informative asides and I think they make the preview. I can see why Dave is worried, that type of thing is rare, but I still think there is plenty of room for improvement, like fps/time, fillrate graphs, info from the questions sent off to nVidia, clarifying my recent confusion about gamma-corrected AA since I can't find a mention of it now for the GF FX (I'm always a selfish b*st*rd when it comes to my expectations from Wavey)...but it's quite a tall order.

 

[Fuz]

Amazing to think that ATi could release a "dustbuster-equipped" 9700 Pro right now that would take the performance crown back immediately, possibly by a considerable margin.

That sums up the situation quite well. Personally, I was expecting much more from the FX.

I was hoping for a new AA method that would perform considerably better than the R300, even with its bandwidth deficit. Thought that might be its trump card.

What about the image quality? From the limited screen shots available, it seems ATI has the edge in AA quality, not to mention 16xAF. The highest image quality available on a FX seems to be slightly worse than whats available on a 9700, and the performance at the highest level is not even comparable, if Hardocp is to be believed.

What I want to see is alot more screen shots of 4xAA running on both cards. From what I saw at Hardocp, ATI's 4xAA seems to do a better job at removing jaggies than the FX at 4xAA.

I hope Beyond3d's review has more info on image quality.

 

[Mintmaster]

What I want to see is alot more screen shots of 4xAA running on both cards. From what I saw at Hardocp, ATI's 4xAA seems to do a better job at removing jaggies than the FX at 4xAA.

It's pretty clear that rotated/jittered grid sub-sampling is no gimmick. When I saw the Beyond3D article on 3DLabs SuperScene AA, I was really shocked that the difference was so great, but when you think about it, it makes a lot of sense.

I don't know why NVidia ignored this feature. I think the IQ argument for ATI is stronger than ever before. You can see why ATI decided to max out with 6x MSAA, because that seems to outclass anything out there, including Matrox's 16xFAA and NVidia's 6xS/8xS.

 

[Brent]

It was highly rushed, we had 2 days to review the card before the NDA lifted.

Believe me we have much more planned for it, I want to get in cpu scaling, cross platform testing, see how much of a performance hit it takes at AA only and then AF only, and more IQ and then just pure gaming with this card to really get a feel for it...

Much more is planned, just didn't have enough time initially...

but initial performance cannot be denied thats for sure... and I can already see AA looks better on the 9700 Pro then the GFFX

 

[Entropy]

Both the review at [H] and at anandtech were a cut above and beyond the expected IMHO. Kudos to them.
Could this be a trend? Can there be hope?
Well, at least there's coffee. * sip! *

Entropy

 

[Fuz]

It was highly rushed, we had 2 days to review the card before the NDA lifted.

-snip-

Much more is planned, just didn't have enough time initially...

but initial performance cannot be denied thats for sure... and I can already see AA looks better on the 9700 Pro then the GFFX

Hey Brent, well done on the preview. I look forward to what you guys have planned. To be honest, I didn't expect much from you guys at Hardocp 8) Hope you guys keep up the good work.

Please give us more on IQ. I think both these cards perform so damn similar, its the IQ we should be comparing.

 

[Nebuchadnezzar]

LOL! From the Hexus.net review. Look at the bars, the 9700 seems to be half as fast as the FX, now look at the numbers

 

[mr]

Hi brent,

It was highly rushed, we had 2 days to review the card before the NDA lifted.

For this short period of time you did a pretty good job.

Much more is planned, just didn't have enough time initially...

Keep up the good work/quality, it's much appreciated.

 

[Joe DeFuria]

Please give us more on IQ. I think both these cards perform so damn similar, its the IQ we should be comparing.

Agreed. (Also agreed on the quality of the review, particularly with the short lead time.)

In the end, I'm willing to bet that one would be extremely hard pressed, if sitting down beside two computers, to tell which one is "faster." In most cases, the FPS numbers are close enough to be considerd "essentially the same."

So in the end, it's going to come down to other factors to really differntiate the two products: image quality, compatibility / stability, price, and other "things" like noise level.

And based on the p/reviews, in every single other aspect than performance (which is a wash overall), the Radeon comes out on top.

From the limited data across all sites so far, there does seem to be a trend that the Radeon fares better in Pentium systems, and the GeForce on AMD. Would be interesting for a single review to look at that....

 

[Colourless]

LOL! From the Hexus.net review. Look at the bars, the 9700 seems to be half as fast as the FX, now look at the numbers

That is what happens if you let Excel decide what limits to use. IMO It's rather sloppy of a site to let that happen, since if people aren't reading the numbers, they get the wrong idea. It's great though for creating false impressions about something... not that I'm saying anything....

 

[Doomtrooper]

Gotta love that graph scaling

 

[Fuz]

According to Hexus.net, the FX has higher image quality than the 9700, yet provides us with no evidence what so ever. I realize that they didn't have much time with the card, but why even mention IQ if they can't back it up with atleast some screen shots.

Give me more IQ comparisons damn it.

 

[K.I.L.E.R]

LOL! From the Hexus.net review. Look at the bars, the 9700 seems to be half as fast as the FX, now look at the numbers

That is what happens if you let Excel decide what limits to use. IMO It's rather sloppy of a site to let that happen, since if people aren't reading the numbers, they get the wrong idea. It's great though for creating false impressions about something... not that I'm saying anything....

Just because lazy bucks and/or idiots don't look at the numbers doesn't mean I don't.
17.6 fps difference is not much at all. In fact I can prove that my Voodoo 3 runs faster than my brothers Geforce 3 by making a "Kruno" graph with no numbers. In other words, a reverse graph. Lower framerates mean a higher pyramid.

 

[Doomtrooper]

According to Hexus.net, the FX has higher image quality than the 9700, yet provides us with no evidence what so ever. I realize that they didn't have much time with the card, but why even mention IQ if they can't back it up with atleast some screen shots.

Give me more IQ comparisons damn it.

Why would you need more screen shots from them ?? Brent posted all the screenshots needed to make your own judgements, as has Anand.