FX vs 9800

[OneOfTheseDays]

yea the FX is simply outclassed. Its loud as heck, takes up an extra slot and does not perform better than the 9800. IF you were smart you'd get yourself a 9700 pro now and upgrade in the future. The 9700's can be cooled with a zalman HSF btw, and you cant have a silent card (or u blow a small 80mm fan on it if your unsure).

 

[dguy6789]

Originally posted by: Rollo
Dguy:
"And no one with these cards will play without AA/AF"
Gee. I guess I'm no one. Gee. It's amazing how you know how everyone uses their VGA? How did you accomplish this? Or are you just talking out of your a$$, because you really don't know 10 people with these cards, let alone everyone?

Question, Are you gonna spend 400 bucks on a card only to have it look like that geforce 3 you got a year and a half ago?? didnt think so. Im not sayng all, just most people, these cards are made to look absolutely gorgeous while have high frames, if you arent goona run AA/AF, then get a radeon 9000 or 8500, they both will do all games highest settings 1280X1024 just fine and dandy without AA/AF. THe 9700+ boards are made to do both. SO why would you spend $400 on a card and do what a $80 card can do? your wasting your timne and money doing that, Period.

 

[touchmyichi]

Comparing the FX to the 9800 pro is a laugh. 9800 pro is a much better card. The FX is a really overclocked NV30. Since Nvidia did not have time to change the architecture of the NV30 in order to compete w/ the 9700 pro they had to attach an extravagant cooling system in order to give it an overclocking boost. Imagine if ATI did the same thing, the difference would be even greater. NV30 was not prepared to deal w/ the 9700 and 9800 pro, even though it is awesome at lower resolutions w/ no AF or AA, but whats the point of a 400 video card if u can't max everything out? The FX was simply released to stay competitive to ATI, which it does. So its not a total failure. Good luck to nvidia on NV35, don't forget what an awesome company this is just because of one poor release.

 

[nRollo]

Lonyo:
"And if someone has the money for one of those cards, they probably want to max out all the options, so I'd say his assumption is fair.
Can you prove it isn't?"
Yes. I have the money for those cards, I have a 9700Pro, and I use 4X AF, period.

"You might want to notice he says "almost every area","
You might want to notice he says "wastes the FX in almost every area"- total bullsh*t, not even CLOSE to true.
Let's see, since I apparently have to read for you:
Pretty even here
9800/5800: 168 vs169, 121 vs129, 84 vs93 The 9800 loses all three, but performance ~ equal

No "wasting" going on here either
9800/5800: 71/69, 70/68, 59/62 9800 wins 2/3, but again, performance is equal

Again, just about equal
9800/5800: 49/49, 43/43, 38/35 Gee, where's all the "wasting"?

You'd have to be pretty dumb to consider this review as showing the FX as "wasted in almost every area"
9800/5800: 142/139, 91/90, 67/64

More of the same
9800/5800: 89/101, 76/87, 66/69

5800 wins again
9800/5800: 53/64, 49/58, 42/46

Let me help you count/evaluate: 9800 wins 7, 5800 wins 10, they tie in 1. A lot more importantly, notice how the 9800s wins are all between 1-3fps? Is that "wasting" to you? The only wins even close to significant are by the 5800.

Synthetic:
9800 wins all three, but by less than 10%

Tied in all three

Obviously at this point the 5800 has proved itself the winner, but again, there's no real significant margin.


Let's move on to the 9800s strong points AA/AF:

9800 wins here, still not what I'd call "wasting"
9800/5800 4xAA 132/117, 88/76, 59/46 At least there's what I'd call our only big difference here at 16X12

AF is the ONLY situation where you can truly say the 9800 "wastes" the 5800
Not much to say here but "well done ATI".

Combining AA/AF can't help the 5800, when it's already lost at both

Well there you have it: The "most" situations the 9800 win in are AA/AF situations, the 5800 wins "most" situations without. If you look at Anand's article, "most" benchmarks are done AA/AF, so it looks like the 9800 wins "most" of the time.

I say the AA/AF reviews are unrealistic. Why are all the AA settings 4X, and the AF settings 8X? What about 2x AA? What about QuincunX AA? What about 2, 4X AF? I think the cards perform so close at these settings they don't bother to publish the results, because what's the point if there's no difference?

"There was a big thread about it in gen hardware, people saying how they don't really like it. Most of the posts were negative, so I'm thinking that most people don't really like/trust THG.
I said anandtech is probably more reliable, and I expect most people woudl agree"
You don't understand statistics at all do you? Or the word "most"? Here you go: One thread on one board does not constitute "most" people. It's a incredibly tiny sample with no statistical power from an incredibly large population. The conclusions you are drawing are pointless. If you walked into a bar tonite and saw 8/10 people drinking Miller, you can't say most people like Miller beer best. Why? Because there are MILLIONS of people drinking beer in bars tonite, and you don't know what other factors affected the choice of Miller at your bar. Similarly, don't you think it makes sense people posting on Anandtech may favor Anandtech?

"Your attitude is not representative of quite a lot of people, "
How many people do you know the attitudes of? Quite a lot? I doubt it.

"you seem a bit of a hot-headed a$$ when it comes to the FX, do you have some kind of interest in it?"
Just tired of all the fan-boy bs in regard to the 9x00 series

"most people woudl agree with is that the 9800PRO is a better card than the FX 5800 "
There you go again, speaking for "most" people.

 

[nRollo]

"Comparing the FX to the 9800 pro is a laugh"
I guess that depends how you use your card. I'm looking for high fps at 10X7 or 12X10, 4X AF. Do you think you can show me benchmark that shows the 9800 having a clear advantage the way I use a card?

 

[SunnyD]

You know why THG isn't trusted?

Simply one and only one thing... THG is PAID but respective manufacturers to get favorable reviews. Used to be the case that they did legit reviews and such, as was the days where THG were posting benches of AMD CPU's trouncing Intel left and right... but then, miraculously, benches seemed to get narrower and narrower, with the respecive benchmark platform being skewed in favor of a certain manufacturer, and over time, that manufacturer would win in rereviews of the same product. Ironic huh?

Tom used to say Rambus sucked balls, I do recall a few months ago a statement from him to the effect of "DDR sucks, Rambus provides a much better solution, but unfortunately is religated to the server market now..."

And it keeps happening over and over, which is why most knowledgable people don't read THG, and do read MULTIPLE other sources and average things out.

 

[Bopple]

Some saying high frams comparisons like 150fps vs 135fps is no matter in real gaming at all.
That's not wrong. But...Then you should buy 135fps one? Or claiming 150fps = 135fps exactly?
Especially when the 150fps one is quieter, more available, cooler, with more IQ, with more matured driver and so on?
And 135fps one has a significant defect with DX9 shading?
(IQ=image qulity)

And about AA/AF, you need to standardized method to compare those between NV30 and R3xx.
For nvidia and ati use different setup of performance setting, you cannot compare the score directly.
With so many reviews-which i don't want to link here. it's so easy to find those. so go for it yourself- we got the next conclusions:

For AA>>
IQwise:
R3xx 6xAA > NV30 any kind of mode(except for 16xAA on which we don't have any objective comparisons yet)
(Actually even R3xx 4xAA is a little bit better than NV30 8xsAA)
Frame-hit-ratio-wise:
They all are on the similar fps scale respectively dependent upon the setup.

And why most benches use 4xAA? Because it is the most popular setup ppl want to use with these kind of cards.

For AF>>
IQwise:
NV30 application = R3xx quality > NV30 balanced > R3xx performance >> NV30 aggressive
Frame-hit-ratio-wise:
NV30 application ~ R3xx quality > NV30 balanced > R3xx performance >> NV30 aggressive

As we see, the only comparable setup is NV30 application vs R3xx quality.
And most of the NV30 winning benches did NV30 aggressive vs R3xx quality. How convenient.

 

[deepinya]

Why spend that much money on a card and NOT max it out? Wouldnt you want the smoothest picture your money can afford?
Those that are arguing about spending this much cash and not using it to its fullest capabilities are just ignorant. Its one thing to love Nvidia but its another to argue for your card with ignorance. I used to be an ATI hater but the facts are clear and the Radeon is the way to go.

Just give ATI its well deserved props and wait for Nvidia to release its next big thing. If they have a card that flat out ownz the 9800 Ill be singing its praises.

 

[touchmyichi]

To all of you defenders I have a question.

If someone handed you an FX and a 9800 pro and said u could only pick one which would you pick?

Thats what I thought.

 

[BenSkywalker]

For GF4/GFFX users, AF is considered a luxury. It is slow, and however Nvidia implements it, it just doesn't look anywhere near as good as ATis.

Actually, that is not the truth. The FX doesn't use as agressive a LOD bias as the R300 boards, which is what Anand looked at when comparing the image quality of anisotropic(a joke), another comical aspect is his using bilinear filtering and claiming it is close to the equal of the trilinear modes(ATi performance setting is simply very poor, much like nVidia's performance mode- quality mode is the only one that should be considered if anyone is looking for quality). Compaing the quality settings both nV and ATi exhibit comparable levels of filtering, with ATi utilizing a more agressive LOD which helps it out on certain angles while introducing texture aliasing on others. You can make the nV boards look the same by setting the LOD to around -0.5 yourself, although certain angles will look better on nV with others looking better on the R300.

The Application mode of the FX is superior to any of ATi's settings, however its performance is horrible. The GeForce4, in terms of quality, kicks them all in the nuts, too bad its so slow in comparison.

This is looking at actual anisotropic filtering, not LOD bias which unfortunately several sites are using in their comparisons. I'm not sure if money changed hands for this to happen or not, but they have been pointed out to the reviewers and they fail to acknowledge it. Digit-Life has a good comparison up of the actual anisotropic filtering, not a paid advertisement for ATi or nVidia which many of the sites seem to be running, and it displays quite nicely that when comparing the R300 AF to the NV30s overall they are fairly close. ATi has the edge on certain angles while nVidia does on others. ATi however sets their LOD bias for their optimal angles leaving less then optimal angles with easily noticeable texture aliasing.

Now, if you want to talk about where the R300 boards have a real edge, use AA

 

[chizow]

Originally posted by: touchmyichi
To all of you defenders I have a question.

If someone handed you an FX and a 9800 pro and said u could only pick one which would you pick?

I'd take the FX. I'm not kidding. I already have a 9700pro softmodded to a 9800pro, and while its easily the highest performing video card I've ever owned, there's been way too many issues with it for me to be completely satisfied and declare it to be "Silicon Jesus" like the rest of you. I was ecstatic when I first installed my Voodoo2 and even more thrilled with my Voodoo 5 when I purchased it; I was also more impressed with my GF3 when I purchased it (I left out the other half dozen or so cards I've owned b/c they weren't as memorable). The Rage 128 and Radeon 8500 are easily the two worst cards I've ever owned.

Driver and hardware compatibility issues with the 9700pro have gradually been reduced with each driver revision, but I'm hardly doing cart-wheels given my expectations of a $315 card and the claims of ATi's driver and compatibility advancements from all the faithful here. To be honest, I feel like I've been lied to.

Thats what I thought.

I guess you thought wrong.

Chiz

 

[Shamrock]

Originally posted by: touchmyichi
To all of you defenders I have a question.

If someone handed you an FX and a 9800 pro and said u could only pick one which would you pick?

Thats what I thought.

I would pick the GF FX 5800 Ultra as well, while the 9800 Pro is faster, I see no real difference in the games I play...before you go off on me and flame me, lemme explain....I race HEAVILY, and I play first person shooters with VERY high intense action....I do no NEED AA/AF, everything is moving so fast you dont have time to stop and enjoy the Niagra falls (eye candy, folks)...and one more thing.

I just could not STAND the BIG GRAY lines going up and down my screen while I am racing/playing it would be too distracting...upto 30% of ATI owners have this problem. That is my main reason I havent upgraded to an ATI card, for fear of those big gray lines.

Anand Poll about gray flickering

And for "everyone" saying "everyone would / wouldnt" I can think for myself, just as Rollo is trying to point out....

touchmyichi my answer is...

GeForce FX 5800 Ultra

Let me also add that you ATI zealots...talking about ATI delivering on their promises...

ATI R400 DELAYED till 1stQ 2004!

Oh, and before you say I am biased, I have owned both NVidia and Ati....I prefer NVidia BY far, even if it means a performance hit.

 

[nRollo]

"Simply one and only one thing... THG is PAID but respective manufacturers to get favorable reviews"
Errr, how do you know this? Links to something other than rumor on a bbs?

 

[nRollo]

"If someone handed you an FX and a 9800 pro and said u could only pick one which would you pick?"

FX Ultra. I have a 9700Pro, it's a nice card and all, but I don't use AA and only 4X AF sometimes. I only play fps games online, and don't think I would notice a difference between the two.

"Thats what I thought"
Bzzzzzt! Wrong answer!

"
The FX doesn't use as agressive a LOD bias as the R300 boards, which is what Anand looked at when comparing the image quality of anisotropic(a joke), another comical aspect is his using bilinear filtering and claiming it is close to the equal of the trilinear modes(ATi performance setting is simply very poor, much like nVidia's performance mode- quality mode is the only one that should be considered if anyone is looking for quality). Compaing the quality settings both nV and ATi exhibit comparable levels of filtering, with ATi utilizing a more agressive LOD which helps it out on certain angles while introducing texture aliasing on others"

I have read about this as well, it's like the old V5 vs GF2 days- yes the GF2 was faster, but the IQ was worse.

 

[SunnyD]

Originally posted by: Rollo
"Simply one and only one thing... THG is PAID but respective manufacturers to get favorable reviews"
Errr, how do you know this? Links to something other than rumor on a bbs?

I'll try to find the articles for you.

Either way, basing your assumptions off of one and only one reviewer's material is just plain dumb.

 

[nRollo]

"Either way, basing your assumptions off of one and only one reviewer's material is just plain dumb"
Can't argue about that. I also can't find non AA/AF benchmarks that show much performance difference between the 9800/Ultra?

 

[ASIMOS]

There are 2 modes to compare 9800 with FX:
1)9800 24FP (96) bit color (standard mode) FX 16FP (64) bit color (9800 has a slight edge in image quality and calculation precision)
Non antialiasing-anisotropic filtering tests SAME SPEED
Antialiasing-anisotropic filtering tests 9800 has much better performance
2)9800 24FP (96) bit color (standard mode) FX 32FP (128) bit color (FX has a slight edge in image quality and calculation precision)
Non antialiasing-anisotropic filtering tests 9800 is twice as fast as FX
Antialiasing-anisotropic filtering tests FX what?
I like to point that Microsoft approved drivers for DX9 specifications doesn?t accept Nvidia FX 16FP (64) bit color.
I also like to point that the vast majority of game developers will support Nvidia FX 16FP (64) bit color through custom modes in their engines.

 

[ASIMOS]

There are 2 modes to compare 9800 with FX:
1)9800 24FP (96) bit color (standard mode) FX 16FP (64) bit color (9800 has a slight edge in image quality and calculation precision)
Non antialiasing-anisotropic filtering tests SAME SPEED
Antialiasing-anisotropic filtering tests 9800 has much better performance
2)9800 24FP (96) bit color (standard mode) FX 32FP (128) bit color (FX has a slight edge in image quality and calculation precision)
Non antialiasing-anisotropic filtering tests 9800 is twice as fast as FX
Antialiasing-anisotropic filtering tests FX what?
I like to point that Microsoft approved drivers for DX9 specifications doesn?t accept Nvidia FX 16FP (64) bit color.
I also like to point that the vast majority of game developers will support Nvidia FX 16FP (64) bit color through custom modes in their engines.

 

[sash1]

Originally posted by: Rollo
"Either way, basing your assumptions off of one and only one reviewer's material is just plain dumb"
Can't argue about that. I also can't find non AA/AF benchmarks that show much performance difference between the 9800/Ultra?

With 2xAA the 5200Ultra absolutely slaughters the 8500 and 9000Pro: Link
With 4xAA, again, 5200Ultra absolutely slaughters the 8500 and 9000Pro: link
Same deal with 2xAA and 4xAF: link
Serious Sam: Almost a close competition (if you take away 20fps from the 5200): Link
And in UT2003, 5200 slaughters the rest: link

On to the 5800:

It pretty much smacks around the 9700 in Quake III: link
And with 4xAA, it's pretty much even: Link
With AA/AF, as res goes up, 9700 pulls ahead. But if you play at 16x12 w/ 4xAA, you have problems: link
UT2003 with AA/AF, same deal as above: Link

~Aunix

 

[nRollo]

Aunix:
You lie. Everyone knows that every ATI card "wastes", "blows away", and "annihilates" every "POS nVidia garbage out there"! It is the Curse of 3dfx!
Don't you read the forums?!

 

[giocopiano]

Originally posted by: chizow

I'd take the FX. I'm not kidding. I already have a 9700pro softmodded to a 9800pro, and while its easily the highest performing video card I've ever owned, there's been way too many issues with it for me to be completely satisfied and declare it to be "Silicon Jesus" like the rest of you. I was ecstatic when I first installed my Voodoo2 and even more thrilled with my Voodoo 5 when I purchased it; I was also more impressed with my GF3 when I purchased it (I left out the other half dozen or so cards I've owned b/c they weren't as memorable). The Rage 128 and Radeon 8500 are easily the two worst cards I've ever owned.

Driver and hardware compatibility issues with the 9700pro have gradually been reduced with each driver revision, but I'm hardly doing cart-wheels given my expectations of a $315 card and the claims of ATi's driver and compatibility advancements from all the faithful here. To be honest, I feel like I've been lied to.

Damn, that doesn't surprise me, but still I'm disappointed. Similar feelings towards V5 and GF3 too. I was hoping someone finally nailed AA, just too bad a leopard never changes its spots.

 

[Ketchup]

Originally posted by: touchmyichi
Comparing the FX to the 9800 pro is a laugh. 9800 pro is a much better card. The FX is a really overclocked NV30. Since Nvidia did not have time to change the architecture of the NV30 in order to compete w/ the 9700 pro they had to attach an extravagant cooling system in order to give it an overclocking boost. Imagine if ATI did the same thing, the difference would be even greater. NV30 was not prepared to deal w/ the 9700 and 9800 pro, even though it is awesome at lower resolutions w/ no AF or AA, but whats the point of a 400 video card if u can't max everything out? The FX was simply released to stay competitive to ATI, which it does. So its not a total failure. Good luck to nvidia on NV35, don't forget what an awesome company this is just because of one poor release.

Wow, that is the biggest bunch of B.S. I have heard in a while. Where do you get off saying the FX has ANYTHING in common with the Geforce 4? Why don't you do a little READING about the card. Actaully, you probably are just confused, because the 9800 PRO is just a 9700 PRO with a higher clock speed. Clock for clock, these cards are pretty much the same speed.

As you crank the resolution, the FX is the faster card. As you crank up, the AF/AA, the ATI is the faster card. So, your choice would depend on which way you want to go to increase visualy quality, Except for the fact that the FX Ultra is a heck of a loud card, which will no doubt turn away some. And then of course there is brand loyalty, which is also going to influence people's decisions.

 

[SSXeon5]

Originally posted by: Rollo
"If someone handed you an FX and a 9800 pro and said u could only pick one which would you pick?"

FX Ultra. I have a 9700Pro, it's a nice card and all, but I don't use AA and only 4X AF sometimes. I only play fps games online, and don't think I would notice a difference between the two.

"Thats what I thought"
Bzzzzzt! Wrong answer!

"
The FX doesn't use as agressive a LOD bias as the R300 boards, which is what Anand looked at when comparing the image quality of anisotropic(a joke), another comical aspect is his using bilinear filtering and claiming it is close to the equal of the trilinear modes(ATi performance setting is simply very poor, much like nVidia's performance mode- quality mode is the only one that should be considered if anyone is looking for quality). Compaing the quality settings both nV and ATi exhibit comparable levels of filtering, with ATi utilizing a more agressive LOD which helps it out on certain angles while introducing texture aliasing on others"

I have read about this as well, it's like the old V5 vs GF2 days- yes the GF2 was faster, but the IQ was worse.

Where is your 9700 pro? So your going to say the FX is better, fan boy your way out of this! You HAD a 9700 pro and how would you notice the difference between the two, have you used the 5800 Ultra? I havent and I can say from what i have read the 9800 or even the 9700 is on my list. I have been through lots of cards, radeon 32mb pci, radeon 64mb, radeon 8500, radeon 9500, geforce GTS, geforce ULTRA (friends), geforce 4 4400 128mb, and i can say ati is by far better. IMHO image quality is alot of it, why would you want 300fps in quake 3 if it looks like SH1T!? And yes I still play it, i love my $160 9500 128mb, its great image quality, fast aa/af, and it isnt load as all hell.

And FU*K toms, my god you cant get the picture, look at anandtechs review they also used the faster setup (3.066GHz/rambus) then that 2700+. THG has never given a solid review and after last year I cant trust them, i just cant. There reviews are bias sometimes and really for any hardware review site that is not a trait to hold. I wouldnt touch a FX with a 10 foot pole, the heat/loadness/power consuming/price is NOT worth it. Im going to sell my 9500 and get a hercules 9700 pro and thats the best bang for the buck for 9700 imho

All in all we need to stop flaming each other, when i say IMHO thats my opinion and thats all, take all of it with a grain of salt. Rollo thats cool you like teh FX man, get one and so be it, thease flame wars arnt really needed. Good day

SSXeon

 

[nRollo]

"Where is your 9700 pro? So your going to say the FX is better, fan boy your way out of this! You HAD a 9700 pro and how would you notice the difference between the two, have you used the 5800 Ultra?"
1. I don't have a 5800Ultra? 2. My 9700 Pro is sitting in the computer behind me until the contractors finish my basement remodelling this week and I get a new desk?

"I have been through lots of cards, radeon 32mb pci, radeon 64mb, radeon 8500, radeon 9500, geforce GTS, geforce ULTRA (friends), geforce 4 4400 128mb, and i can say ati is by far better."
I've been through many more cards than that in my 14 years of computer gaming, but have a soft spot for ATI IQ as well.

"And FU*K toms, my god you cant get the picture, "
Calm down, my reading a popular web page isn't worth you having a heart attack over...

"There reviews are bias sometimes and really for any hardware review site that is not a trait to hold"
Nobody in business is completely objective. Do you think it's a coincidence that this site changed their whole review focus* to emphasize one of their largest advertiser's selling points? That's just the way business is. (*By this I mean the FX Ultra review was the first to not have a balance of non AA/AF and AA/AF benchmarks, with a statement "Since the focus of this review is the performance of the two fastest desktop GPUs there are a few assumptions we can make, the biggest being that you don't buy a card like the GeForce FX 5800 Ultra to run without Anti-Aliasing or Anisotropic Filtering enabled" Obviously there are some of us speed junkies left who DO buy cards like this and run them w/o these features)

"I wouldnt touch a FX with a 10 foot pole, the heat/loadness/power consuming/price is NOT worth it. "
No one will force you to buy an FX. I promise.

"All in all we need to stop flaming each other"
Err, your whole post is a flame? You swear, yell, call me stupid, and pretty much demand everyone share your viewpoint? Practice what you preach?

 

[oldfart]

Except for the fact that the FX Ultra is a heck of a loud card, which will no doubt turn away some

That made it DOA for me. People who like quiet PC's will not even look at it. It doesn't matter how good/bad it performs. I saw a guy on this forum put his on EBay today because of the noise. BTW, current card is Ti4600, GF3 before that, ATI Radeon 64 original before that one. No brand loyalty here either. Next card is a 9700 Pro which will be here next week.