Toms look at the Geforce FX

[Tompa]

hmmmm....how come these guys come to different conclusion than the other sites....maybe someone more educated than me can make some comments....

Quote from the conclusion from Toms hardware:

" NVIDIA takes the crown! No question about it - the GeForceFX 5800 Ultra is faster than the competition from ATI's Radeon 9700 PRO in the majority of the benchmarks. However, its lead is only slight, especially compared to the distance that ATI put between its Radeon 9700 PRO and the Ti 4600. Still, when compared to its predecessor, the GeForce4 Ti, the FX represents a giant step forward.

The GeForceFX 5800 Ultra is irrefutably the fastest card in most of the tests - but at what price? The power of the FX relies on high clock speeds, which in turn require high voltages and produce an enormous amount of heat. The consequence is that extensive (and expensive) cooling is necessary. Add to that the DDR-II memory, the price of which is quite high, due to the small production numbers. Even the 12-layer board layout is complex and expensive.

It will be difficult for NVIDIA to push its GeForceFX 5800 Ultra. Radeon 9700 PRO cards are only slightly slower, and, because they've been out on the market for months now, they're much less expensive. Also, because they deliver 3D performance with much slower clock speeds, they do not require extensive cooling - and that's nice for your pocketbook as well as your ears.

Still, despite expectations to the contrary, the official price for the FX 5800 is $399 plus tax and that seems pretty aggressive and attractive. This makes it identical to the launch price of the GeForce4 Ti4600 and the ATI Radeon 9700 Pro. The "normal" version of the 5800 will be somewhat less expensive. It's surprising that the GeForceFX GPU, clocked at 500 MHz, only gains a small lead over the R300 GPU (VPU), which is modestly clocked at 325 MHz in comparison.

It remains to be seen how long NVIDIA, with its FX 5800, can maintain a lead over ATI. ATI has already started to hint at a faster-clocked R350 to come in the next weeks (according to rumor, it will have a 400-425 MHz core and 800 MHz memory).

Nevertheless, enthusiasts will, without a doubt, love the GeForceFX 5800 Ultra. It is a monster card! And it has a look that is similarly spectacular to the 3dfx Voodoo5 6000 at the time of its launch. "

...end quote....


// Tompa

 

[bridpop]

ATI is on mature drivers with the r300, whereas the Geforce FX is late, disasterously launched, and the drivers appear to be pretty much sucking.....

.... However, give it 6 months, with a few driver revisions and the usual "Detonator 4/5/6" release, you'll see that card pulling ahead in leaps and bounds ahead of the R300 and 350.... everyone keeps mentioning the new ATI product coming out in march, but remember we have the NV31/34/35 just around the corner too.

If theres one thing Nvidia is good at, its squeezing as much performance as possible out of its cards... in a few months it will be killing everything in its path...

but yeah.... at the moment ... it pretty much sucks.... who wants a jet engine fan system and performance more or less the same as my 120 pound 9500 with patched drivers and a little overclocking

Tomshardware can see through the initial dissapointment methinks.... but hey.... it IS the fastest ..... at a price...

 

[Tahir2]

NV31/NV34 will be slower versions for the mid market segment.
The R350 not enough is known about it to see if the NV30 can compete with it. NV35 maybe but not NV30/NV31/NV34.

ATI can improve their drivers too. They haven't squeezed everything out of it and like you said NVIDIA definitely will be squeezing a lot out of the GFFX over the next few months - but by then it is too late. First impressions count.

 

[KimB]

.... However, give it 6 months, with a few driver revisions and the usual "Detonator 4/5/6" release, you'll see that card pulling ahead in leaps and bounds ahead of the R300 and 350.... everyone keeps mentioning the new ATI product coming out in march, but remember we have the NV31/34/35 just around the corner too.

I'd give it one month for the drivers to be noticeably improved, but full optimization is a ways off yet. And the Radeon 9700's drivers really aren't that polished, at least compared to the GeForce4's drivers, so take from that what you want...

 

[Doomtrooper]

Its is not the fastest whatsoever..it loses to a Stock 9700 by a good margin in almost all games with AA and AF (which is one of the selling features of this card)...image quality is superior on the 9700 with its RGMS....

Tomshardware is a joke...I wish that site would just go away..credability it did have is now gone...they are simply misleading their readers and consumers badly.

 

[demalion]

I sort of liked some of the details of the benchmarking, like mentioning triangle throughput for CodeCreatures and mentioning the Z buffer issue.

But, with the rather incredibly blatant nVidia cheerleading going, it makes me wonder if someone else came up with the benchmarking and Lars came along and wrote all the comments.

It is like the conclusion was written before anything else was.

 

[KimB]

Its is not the fastest whatsoever..it loses to a Stock 9700 by a good margin in almost all games with AA and AF (which is one of the selling features of this card)...image quality is superior on the 9700 with its RGMS....

Are you referring to the Anandtech benchmarks? Those were using performance anisotropic, which means bilinear filtering, vs. the FX's trilinear. That's certainly not a fair comparison.

 

[Doomtrooper]

DP

 

[Doomtrooper]

Chalnoth just stop the damage contol..you are smarter than this... Look at the screen shots from Hardocp in UT 2003.


Nahhh I'll post it...

HQ Nvidia..look at the frame counter

HQ ATI

Sorry for the 56K'ers but this is just too obvious.

 

[KimB]

Chalnoth just stop the damage contol..you are smarter than this... Look at the screen shots from Hardocp in UT 2003.

Nahhh I'll post it...

HQ Nvidia..look at the frame counter

So? The xS modes are pointless for today's games. The FX really shouldn't be compared in image quality using them (supersampling is just too slow). Yes, nVidia dropped the ball with antialiasing.

The xS modes are pretty much only good for older games. Let the people who are still addicted to CounterStrike rejoice.

 

[Doomtrooper]

Ok then 4X MS then...

http://www.hardocp.com/article.html?art=NDIxLDQ=

Whats the point ??

 

[antlers]

So? The xS modes are pointless for today's games. The FX really shouldn't be compared in image quality using them (supersampling is just too slow). Yes, nVidia dropped the ball with antialiasing.

The xS modes are pretty much only good for older games. Let the people who are still addicted to CounterStrike rejoice.

The point of the comparison is that on at least one modern game, there is a mode on the 9700 that has both playable frame-rates and visibly better image quality than the highest image quality mode on the FX, even though the FX has unplayable framerates.

Actually, the aliasing on the FX frame is worse at 8xS than I would have imagined. Perhaps there is some sort of image capture problem that's not showing how these really show up on the screen.

 

[Lincoln]

There is something wrong with the GFFX UT 2003 screenshot, it appears that the fog is missing.
I can look at the exact same scene with 8xS on my Ti-4600 and get a substantially better looking image.

These AA benchmarks are obviously not "apples to apples", due to the supersampling component of the "XS" modes.
The problem is, you almost have to use these modes to approach the edge AA quality of the Radeon 9700 in 4x and higher.
Unfortunately it's kind of an "all or nothing" approach, either add supersampling with the "XS" modes and dramatically reduce
performance to get better edge AA (and the good things about SS), or use the lower level AA modes and get relatively crappy
edge AA and none of the benefits supersampling offers.

Given the supposed 3DFX influence, I'm quite puzzled by the AA options the GFFX provides.

Lincoln

 

[LeStoffer]

Given the supposed 3DFX influence, I'm quite puzzled by the AA options the GFFX provides.

Indeed.

 

[Fred]

'it loses to a Stock 9700 by a good margin in almost all games with AA and AF '

Thats really stretching it. From Anands review, at the same settings, we see the Nv30 beating the ATI card in nearly every benchmark. Performance is not the Nv30s problem.

Its the relative quality of ATI's IQ solution thats still so impressive, particularily the antialiasing.

Do you prefer 10% more speed, or 10% better IQ? Most people choose the latter.

 

[OpenGL guy]

There is something wrong with the GFFX UT 2003 screenshot, it appears that the fog is missing.
I can look at the exact same scene with 8xS on my Ti-4600 and get a substantially better looking image.

I noticed the fog problem as well. To me it seems that fog is there (some things are fogged), but some objects seem "under fogged".

These AA benchmarks are obviously not "apples to apples", due to the supersampling component of the "XS" modes.
The problem is, you almost have to use these modes to approach the edge AA quality of the Radeon 9700 in 4x and higher.

The HardOCP screenshots comparing GFFX's 8xS AA w/ 8x AF vs. the 9700 Pro's 6x AA w/ 16x AF show much better edge quality on the 9700 Pro. Even on the thin angled wire towards the upper left, where the ordered grid of 8xS should shine, looks better on the 9700 Pro. Surprising to me.

Another thing I noticed was that there didn't appear to be any extra filtering going on due to the super-sample component in 8xS. Notice the HUD icons look very sharp.

 

[Fred]

I noticed that too. Its a bit mysterious.

All in all, the FX antialiasing quality seems very raw, and buggy. Whether its through hardware failure or Driver issues is as yet to be seen.

Id like to see a few counterstrike pics.

 

[Doomtrooper]

'it loses to a Stock 9700 by a good margin in almost all games with AA and AF '

Thats really stretching it. From Anands review, at the same settings, we see the Nv30 beating the ATI card in nearly every benchmark. Performance is not the Nv30s problem.

Its the relative quality of ATI's IQ solution thats still so impressive, particularily the antialiasing.

Do you prefer 10% more speed, or 10% better IQ? Most people choose the latter.

What does the picture above show ..MAX Detail..who has faster frames

So we have both cards cranked to maximum settings, one card delivering better frames and better IQ...I don't understand your statement at all.
Another point, nobody overclocked the 9700 which was dissappointing, I understand the reason to only overclock the FX to see its limits, yet I think a mildly overclocked 9700 would have shown a even different picture.

 

[Fred]

So you produce one benchmark, with a supersampling mode enabled on the FX, and claim its representative of 'nearly all games with AF and AA'. This despite Anand's fairly comprehensive results.

Nod that makes sense

A better argument would be that while it seems the FX wins with 4*, going to progressively higher antialiasing modes, seems to bode in the 9700's favor, at least the gradient goes their way. Particularily since the #XS modes are necessary in order to compete with the IQ.

 

[Doomtrooper]

Wait a minute..we are talking about MAX details on both cards, you can start a arguement over details on how its implemented on each card but that doesn't bode well for the average consumer. Does he really care Super Sampling is being used , or even know what it is.
I'm also refering to the 4X MS benchmarks, which are either FLAT or the 9700 leading at the Hardocp review...screen shots that show the frame counter.

Either way..with FSAA and AF enabled its even or losing in most benchmarks..