FX 5900XT 128mb VS 9800PRO 128mb

[GeneralGrievous]

Why not just go up to a 6800 GT?

 

[Algere]

Originally posted by: Gamingphreek
But the NV3x series is plagued with problems right now. Also Nvidias drivers aren't nearly as rock solid as they used to be. I prefer Nvidia but ill buy anyone who makes a better card. Will the 9800Pro score better on 3dmark (just out of curiosity).

As is said earlier unless Anyone respond ASAP. Is the 9800PRo a fairly decent jump over my current card, a MSI 5900XT-VTD128. I can do without the VIVO i suppose but im looking at the MSI 9800PRO. Also does anyone think that if i RMA Msi and ask for the 9800Pro and that i pay the extra 10$ do you think they will let me switch.

Depends, The 5900 "might"(not sure) score higher on 3DMark03 than 9800 Pro and the ASUS 9800 Pro's come with VIVO so that's something you won't have to be without.

The jump between the 5900 and 9800 Pro isn't much except in PS 2. Is it worth it to switch? well that's entirely up to you.

Final note, I'd doubt they'd let you switch but I suppose it's worth a try.

 

[Gamingphreek]

Im not looking to spend more than 20$ in the switch. I think the Asus board is well over 220. Im not too worried about 3dmarks as long as it performs better in real games. Will i see any difference in framerate in games like halo and far cry and doom III ya know the usual.

I can get an RMA... its a very long stretch but on the RMA reason it has video noises. Well my monitor is a 19" and its getting quite old almost 10 years now so it cant support 85Hz at 1280x1024 and that stuff. But thats besides the point it i suppose has bad capacitors in it and it whines loudly. So much that i have to walk away lol. So ill stretch the truth lol and use that as my excuse.

I already know what everyone is gonna say... "sounds like you need a new monitor man"... yeah but until this one dies im fine ill just lower the refresh rate and that sh!t.

 

[BFG10K]

I'll just keep on pointing out the gross exaggerations in your posts

No, you'll continue to talk utter rubbish using grossly inaccurate generalizations, all the while ignoring the facts in front of you.

(e.g. how you love to use only the 16X12 4X8X bench as proof of your position when neither card can run the game at anything close to playable framerates at those settings)

Nobody needs 32 bit colour. -3dfx.
Nobody needs high resolution, AA, AF or shaders. -Rollo

And let's play along for a minute: if the cards are "equal" at the CPU limited settings that you run at, how did you arrive at the conclusion that the card that runs the slowest at high resolutions and/or with shaders should be the preferred option?

Do you have even the slightest inkling of how ludicrous your arguments are? Pointing out your stupidity is not the same as restricting choice no matter how much you scream that it is.

 

[nemesismk2]

Originally posted by: Gamingphreek
But the NV3x series is plagued with problems right now. Also Nvidias drivers aren't nearly as rock solid as they used to be. I prefer Nvidia but ill buy anyone who makes a better card. Will the 9800Pro score better on 3dmark (just out of curiosity).

As is said earlier unless Anyone respond ASAP. Is the 9800PRo a fairly decent jump over my current card, a MSI 5900XT-VTD128. I can do without the VIVO i suppose but im looking at the MSI 9800PRO. Also does anyone think that if i RMA Msi and ask for the 9800Pro and that i pay the extra 10$ do you think they will let me switch.

I have a MSI 5900 XT VTD128 just like you and I also own a 9800 Pro, personally I think it would be a waste of money sidegrading from a 5900 XT to a 9800 Pro. MSI still sell the 5900 XT so I think it is unlikely they will give you a different video card, imo they will just send you a replacement 5900 XT like you already own! 🙂

Here is some results with my xp3200 and the two video card you are interested in:- (all game benchmarks at 1280x1024, 8x4 scores in brackets and with sound enabled ofcourse!)

3dmark 2003

fx5900xt = 5170
r9800 pro = 5860

x2 the threat

fx5900 xt = 89 (67)
r9800 pro = 73 (66)

ut2003

fx5900 xt = 59 (34)
r9800 pro = 55 (36)

halo

fx5900 xt = 37
r9800 pro = 35

 

[nRollo]

Originally posted by: BFG10K

I'll just keep on pointing out the gross exaggerations in your posts

No, you'll continue to talk utter rubbish using grossly inaccurate generalizations, all the while ignoring the facts in front of you.

I don't ignore any "facts". I only evaluate in a more meaningful way then you do. I didn't say everyone needed the "superior" ATI shader performance like you did, a year before any games that used shaders existed. I didn't point at Shady Days benches like you did, look what a joke that turned out to be. I don't look at differences that are irrelevant because the framerates on both cards are unplayable either.

(e.g. how you love to use only the 16X12 4X8X bench as proof of your position when neither card can run the game at anything close to playable framerates at those settings)

Nobody needs 32 bit colour. -3dfx.
Nobody needs high resolution, AA, AF or shaders. -Rollo

And let's play along for a minute: if the cards are "equal" at the CPU limited settings that you run at, how did you arrive at the conclusion that the card that runs the slowest at high resolutions and/or with shaders should be the preferred option?[/quote]
It's irrelevant if the gpu limited settings aren't playable. Whether you can run a game at 20 or 30fps average doesn't matter a whole lot to me, because both suck. (even if one is 50% faster than the other! Jinkies!)
You're right about one thing though- I do think nobody needs hi res, AA, AF if their computer can't keep the minimum framerate above 40-50fps with them on, and no one needs better shaders if the tradeoff is not being able to use your AA/AF, or higher res.

Do you have even the slightest inkling of how ludicrous your arguments are? Pointing out your stupidity is not the same as restricting choice no matter how much you scream that it is.

LOL- I could say the same to you, and more than a few people have agreed with me. "What would you rather ride?!?! A cow or a goat?!?!"

I'll stop pointing out your fan boy behaviour when you stop telling people the only relevant benchmarks are the ones at unplayable framerates, on unreleased games, etc..

I don't really care what you think of my level of intellect. I made it through two bachelors at good colleges and work in IS. I don't care if you have a Ph.D behind your 10K (which I really doubt)- I don't fit any standard definition of stupidity any more than your definitions of product "superiority" fit and standard definition of "common sense".
😉

 

[nRollo]

Originally posted by: nemesismk2

Originally posted by: Gamingphreek
But the NV3x series is plagued with problems right now. Also Nvidias drivers aren't nearly as rock solid as they used to be. I prefer Nvidia but ill buy anyone who makes a better card. Will the 9800Pro score better on 3dmark (just out of curiosity).

As is said earlier unless Anyone respond ASAP. Is the 9800PRo a fairly decent jump over my current card, a MSI 5900XT-VTD128. I can do without the VIVO i suppose but im looking at the MSI 9800PRO. Also does anyone think that if i RMA Msi and ask for the 9800Pro and that i pay the extra 10$ do you think they will let me switch.

I have a MSI 5900 XT VTD128 just like you and I also own a 9800 Pro, personally I think it would be a waste of money sidegrading from a 5900 XT to a 9800 Pro. MSI still sell the 5900 XT so I think it is unlikely they will give you a different video card, imo they will just send you a replacement 5900 XT like you already own! 🙂

Here is some results with my xp3200 and the two video card you are interested in:- (all game benchmarks at 1280x1024, 8x4 scores in brackets and with sound enabled ofcourse!)

3dmark 2003

fx5900xt = 5170
r9800 pro = 5860

x2 the threat

fx5900 xt = 89 (67)
r9800 pro = 73 (66)

ut2003

fx5900 xt = 59 (34)
r9800 pro = 55 (36)

halo

fx5900 xt = 37
r9800 pro = 35

You are clearly a nVidiot Nemesink. By BFGs universal definition of superiority, the 9800Pro is clearly superior [/b] because it wins the UT2004 benchmark by 2fps and 3dmark by 700 whole points!
:roll:

 

[nemesismk2]

Originally posted by: Rollo

Originally posted by: nemesismk2
You are clearly a nVidiot Nemesink. By BFGs universal definition of superiority, the 9800Pro is clearly superior [/b] because it wins the UT2004 benchmark by 2fps and 3dmark by 700 whole points!
:roll:

The only game I've noticed that the 9800 Pro performs much better than my 5900XT is Far Cry which is probably because I only have the demo. However my 2.5Ghz XP2500m is in the PC with my 5900XT so it still manages 67fps (high settings) which I think is pretty good! 🙂

 

[ForceCalibur]

Omg, I can't believe this argument is still going on. It's been decided already which is one is better by the market and consumer preference.

The 9800 PRO IS BETTER THAN THE 5900XT.


But then Ferrari is better than a BMW!
And a Mercedes is better than a Honda!

For the price you pay, the 5900XT and the 9800 Pro gives out similar price/performance, with the 5900XT being better in some cases because of its nearly 50 dollar markdown from the 9800 Pro.

And just let the freken thread die!

And yes, if its gonna cost you 20 bucks to upgrade to 9800 Pro, Why not? I blow 20 bucks like nothing (and it is NOTHING in the long run)

 

[RussianSensation]

Originally posted by: ForceCalibur

And just let the freken thread die!

Amen!

I blow 20 bucks like nothing

I don't know how many people want to hear about your sex life with a 20 dollar bill Force 😀 j/k


Gamingphreek if you play far cry and intend to keep the card for a while longer before upgrading and do want to play HL2 at decent framerates then sure it is worthwhile to switch to the 9800Pro for $20.

 

[nemesismk2]

Originally posted by: RussianSensation

Gamingphreek if you play far cry and intend to keep the card for a while longer before upgrading and do want to play HL2 at decent framerates then sure it is worthwhile to switch to the 9800Pro for $20.

How many fps equals "decent framerates"? While your looking into your crystal ball, will the 9800 Pro deliver your so called "decent framerates" with Doom3, Thief - Deadly Shadows etc as well? 😉

 

[BFG10K]

I don't ignore any "facts".

Yes you do. You ignore GPUs and instead focus on comparing CPUs to each other.

I didn't say everyone needed the "superior" ATI shader performance like you did, a year before any games that used shaders existed.

No, you instead claimed that nobody needed shader performance and that the cards are equal based on your CPU tests.

Your actions or non-actions have absolutely no bearing on the actual situation no matter how much you think they do.

It's irrelevant if the gpu limited settings aren't playable.

So your logic boils down to this:

Card A and card B are the same at low detail settings.
Card A is faster than card B at high detail settings.
==================================
Therefore - using Rollo logic - one should pick card B

Again do you have any idea how idiotic your arguments are?

LOL- I could say the same to you,

No you couldn't.

I'll stop pointing out your fan boy behaviour

So pointing out that your picking of card B above is a stupid decision makes me a fan boy? Tell me then, what does picking card B make you then?

 

[Zebo]

Here it shows graphically how the 6800Ultra gets dominated by ATI sometimes as much as 300% with eye candy turned on. These cards are even until then.

http://img.hexus.net/v2/graphics_cards/ati/r420/images/3dmark.png

http://img.hexus.net/v2/graphics_cards/ati/r420/images/3dmark03.png

http://img.hexus.net/v2/graphics_cards/ati/r420/images/cod.png

http://img.hexus.net/v2/graphics_cards/ati/r420/images/ss2.png

http://img.hexus.net/v2/graphics_cards/ati/r420/images/ut2003.png

http://img.hexus.net/v2/graphics_cards/ati/r420/images/x2.png

 

[rbV5]

Here it shows graphically how the 6800Ultra gets dominated by ATI sometimes as much as 300% with eye candy turned on.

You mean with 8xAA turned on the 6800. SuperSampling makes 6800's 8XAA mode inefficient, so of course any benchmark using 8XAA on the 6800 will show a large performance hit. Its not "eye candy" turned on, its only really showing ATI's 6XAA mode is more efficient than NV's 8XAA mode.

 

[Gamingphreek]

Secondly that graph is extremely ATI biased. I mean look at the numbers. Any other reviews will show you that Nvidia has largly fixed that problem and doesn't suffer AS MUCH. I think you need to look at some other reviews.

-Kevin

 

[nRollo]

So your logic boils down to this:

Card A and card B are the same at low detail settings.
Card A is faster than card B at high detail settings.
==================================
Therefore - using Rollo logic - one should pick card B

Again do you have any idea how idiotic your arguments are?

You forgot the most important parts of my idiotic logic BFG-
1. I don't say you should pick card B if there's a playable setting one card runs at and the other doesn't
2. I usually say there's no compelling reason to buy either card over the other if they are approximately the same at the playable settings.

Unlike you, man of logic, who often states that if the cards are equal at all the playable settings, but cards A is at 20fps at the highest setting and card B is at 10fps, card A is "superior" and "twice as fast".
Apparently unlike you, I know I can't play a game at 10 or 20 fps, so I say either framerate is irrelevant because it's unusable. The only logic I could see in your position is there may be another game where this difference would show as the ability to play at 16X12 instead of 12X10.

No, you instead claimed that nobody needed shader performance and that the cards are equal based on your CPU tests.

For a year and a half you've been yelling about DX9 and Shader performance of the R300 series BFG. Do you even own a DX9 PS2 game?
You're right, I claimed nobody needed good PS2 shader performance for that year and a half, and as there were no games using it, I'd say I was right. Your argument was as "logical" as saying:
"Camaros have an airfoil that can stabilize the ride when cars can fly, Mustangs do not. Cars can't fly yet, but I'll be ready for flying in my Camaro when they do".
Of course, by the time the cars CAN fly, the airfoil technology is so obsolete only unemployed people and kids are stuck with Camaros.

Like I said, call me an idiot all you want. I don't notice you chiming in about your accomplishments in academia or the rocket science work you do, so I'm guessing you don't have much to chime in with as far as your scholarly accomplishments go in terms of how you are qualified to tell me I'm less intelligent than you are.

 

[Maskirovka]

i see both sides of the Rollo/10k argument...

rollo: the cards are basically the same at playable settings
bfg: 9800p is better with AA/AF/high res
rollo: 9800p is unplayable with AA/AF/high res so who cares?

http://www.xbitlabs.com/articles/video/display/gainward-5900xt_5.html

i disagree that CoD, 1280, 4xAA 8xAF at 83avg fps is unplayable...compared to the 67avg the 5950U gives (also not unplayable...but the 5950 is much more expensive). and the 9800's beat the FX's without AA/AF in far cry according to the xbit benches as well.

http://www.firingsquad.com/hardware/ati_radeon_x800/page23.asp

there's also a ~20fps difference between the 9800p and the 5950u at 1024 in this firingsquad far cry bench.

sure it of course comes down to the games you're playing...especially if you're not going to upgrade for a year or so...and whether or not you want to consider that all the games the 9800p is better at are a better reflection of what you're going to be playing in the future.

i can see how a 5900xt would be attractive to a person with a ~$160 budget...but if your budget is more like ~$200, i don't see how the 5900xt is worth it (feel free to say why you think it is...i'm not trying to argue...)

....and if your budget is more like $350, buying a 5950u would be dumb since you could just buy a more budget card now or save for a bit and get an x800pro/xt or 6800u

 

[BFG10K]

1. I don't say you should pick card B if there's a playable setting one card runs at and the other doesn't

The swinging goalpost of your definition of playable is one of the problems with your logic, not to mention that you're constantly ignoring cases where the 9800's extra speed puts itself into the playable range.

2. I usually say there's no compelling reason to buy either card over the other if they are approximately the same at the playable settings.

Which is utter rubbish given the 9800 Pro is both faster and cheaper. If that's the evidence you need to determine that a coin flip is the best method to decide which card you should get then you really have issues that extend beyond the scope of the thread.

Unlike you, man of logic, who often states that if the cards are equal at all the playable settings, but cards A is at 20fps at the highest setting and card B is at 10fps, card A is "superior" and "twice as fast".

Do you understand the dictionary definitions of "superior" and "faster"? Do you speak English? Or is that something else you choose to ignore in order to stay in your dreamworld?

Apparently unlike you, I know I can't play a game at 10 or 20 fps, so I say either framerate is irrelevant because it's unusable.

There is no difference between a glider and a balloon because I'm not in an F-16, eg Rollo?

Likewise, there's no difference between one bedroom unit and a three bedroom house because it doesn't compare to Buckingham Palace?

And there's no difference between a bike and walking because Ferrari is faster than both of them?

For a year and a half you've been yelling about DX9 and Shader performance of the R300 series BFG. Do you even own a DX9 PS2 game?

That's called a strawman.

 

[BFG10K]

i disagree that CoD, 1280, 4xAA 8xAF at 83avg fps is unplayable...compared to the 67avg the 5950U gives (also not unplayable...but the 5950 is much more expensive).

You're really wasting your time with Rollo. He'll ignore anything that doesn't fit into his dreamworld and call anybody a fanboy that uses logic to disprove him.

 

[nRollo]

Originally posted by: Maskirovka
i see both sides of the Rollo/10k argument...

rollo: the cards are basically the same at playable settings
bfg: 9800p is better with AA/AF/high res
rollo: 9800p is unplayable with AA/AF/high res so who cares?

http://www.xbitlabs.com/articles/video/display/gainward-5900xt_5.html

i disagree that CoD, 1280, 4xAA 8xAF at 83avg fps is unplayable...compared to the 67avg the 5950U gives (also not unplayable...but the 5950 is much more expensive). and the 9800's beat the FX's without AA/AF in far cry according to the xbit benches as well.

http://www.firingsquad.com/hardware/ati_radeon_x800/page23.asp

there's also a ~20fps difference between the 9800p and the 5950u at 1024 in this firingsquad far cry bench.

sure it of course comes down to the games you're playing...especially if you're not going to upgrade for a year or so...and whether or not you want to consider that all the games the 9800p is better at are a better reflection of what you're going to be playing in the future.

i can see how a 5900xt would be attractive to a person with a ~$160 budget...but if your budget is more like ~$200, i don't see how the 5900xt is worth it (feel free to say why you think it is...i'm not trying to argue...)

....and if your budget is more like $350, buying a 5950u would be dumb since you could just buy a more budget card now or save for a bit and get an x800pro/xt or 6800u

It is true Maskirova, the 9800Pro is a better card at CoD and Far Cry than the nV35 series, but those are just two games. I can find benchmarks where the 5900 is the better card.

I've heard a game that about to become the only one I play for a while is coming out in two weeks- Doom3. The nV35s may well be faster at that, the nV40s surely will be.

BTW- there are some decent reasons to pay a bit more for a nVidia based card:
1. Support American firm if you're American
2. Get real drivers and performance if you use Linux
3. nVidia drivers/cards set standard for compatibility and stability
4. usually better at OpenGL
5. Much better feature set on nV40 series (and I believe we'll start actually seeing cards in the next week or two)
6. Possible SM3 performance enhancements
7. Not buying rehash of 2 year old chip like X800 series.

 

[GeneralGrievous]

I've heard a game that about to become the only one I play for a while is coming out in two weeks- Doom3. The nV35s may well be faster at that, the nV40s surely will be.

Do you think the NV40s, especially the GT, will be out in time for Doom 3? Unlikely.

How ironic. There's a solid chance that ATI has the best card to play Nvidia's flagship title at its time of release.

 

[nRollo]

Originally posted by: GeneralGrievous

I've heard a game that about to become the only one I play for a while is coming out in two weeks- Doom3. The nV35s may well be faster at that, the nV40s surely will be.

Do you think the NV40s, especially the GT, will be out in time for Doom 3? Unlikely.

How ironic. There's a solid chance that ATI has the best card to play Nvidia's flagship title at its time of release.

For how long General? A week? Two?
Here's some better irony: all the slobs that bought 9600XTs and 9800XTs for the HL2 coupon will have slow, outdated cards by the time HL2 is released?

Besides all this, I read on another forum a guy said the CompUSA in his town will have 6800Us in today, and everything I've read says 6800Us will be out in the next few weeks.
I'd say anyone buying a X800Pro to have the best Doom3 card is "smoking something hallucinogenic"

 

[Corporate Thug]

BTW- there are some decent reasons to pay a bit more for a nVidia based card:

1. Support American firm if you're American

i like to support American companies but that isnt going to make the decision for me. for the same reason, i will probably never buy the an american car over a japanese car

2. Get real drivers and performance if you use Linux

legit point if you are one of the few people using linux

3. nVidia drivers/cards set standard for compatibility and stability

i havent had many problmes with ATi, i wasnt able to run SS: PT but i switched to a nvidia card and had the same problem

4. usually better at OpenGL

legit

5. Much better feature set on nV40 series (and I believe we'll start actually seeing cards in the next week or two)

legit

6. Possible SM3 performance enhancements

not buying into the SM3 hype until i see its merits

7. Not buying rehash of 2 year old chip like X800 series.

i dont understand this. people including nvidia keep beating this point to death but if i'm getting that kind of performance off two year old tech, then more power to ATi.

 

[nRollo]

The rehash of two year old tech is just a personal thing with me. I bought the 9700Pro when it came out for about $400.. I bought the 9800 Pro when it came out for about $400. Now, I can buy the same chip with four more pipes and some faster RAM for, you guessed it, $400.

It just doesn't feel like buying a new card when you're just getting a faster version of the old card and they're taking another $400.. At least the 6800Us are a new design with some new features in them. I don't even care if they are a little slower in some benches, drivers should improve their performance, and it feels like you're getting a new card.

If ATI release the X900 next year with four more pipes am I supposed to give them another $400 just because it benches well? Methinks not. 😉

 

[BFG10K]

It just doesn't feel like buying a new card when you're just getting a faster version of the old card and they're taking another $400..

LOL, this from the guy that purchased a 5800 Ultra and 9800 Pro twice and a 9700 Pro once, each time replacing his previous card with the next.

He has a problem with paying to get faster performance from the X800 series but he has no hesitation to pay to downgrade to a 5800 Ultra - twice.

If ATI release the X900 next year with four more pipes am I supposed to give them another $400 just because it benches well?

No, you should downgrade to a 5800 Ultra for the third time. :roll: