more future proof nvidia 5900 ultra or ati 9800 pro?

[CaiNaM]

snowman, skipping quote feature here, but honestly, 4 month old tests on 6 mo old drivers using a static bench across different cpus in an uncontrolled environment is supposed to provide what insight, exactly, to real world gaming, or what we are playing today and will be playing tomorrow?

do you really expect anyone to debate that?

 

[kylebisme]

i just went there because the results still hold true to more recent testing and wanted to copy/paste the results with the % difference tallys.

here is some recent shadermark tests.

 

[Genx87]

It has been my experience that there is that one game or generation of games that just make everything crumble.
It doesnt matter if you owned the 350 dollar card or the 100 dollar card from that generation. It just makes them all cry.

I wouldnt expect either card to outlast each other much in terms of futureproof. When I buy my hardware I buy it for the now and what it can do today. If I worried about what it can do in 3-6 months or more all I would do is lose my mind because it would fail almost everytime.

this latest round for me has been good. I biult the actual computer last April and got the video card in August. So far it runs everything awesome and honesly dont expect an upgrade until maybe next fall. But who knows. I played the Far Cry demo and had no problems at 1600X1200. If that is all they can throw at us for this year then maybe just maybe Ill have a computer that will last 24 months 🙂

 

[mikevine69]

"this latest round for me has been good. I biult the actual computer last April and got the video card in August. So far it runs everything awesome and honesly dont expect an upgrade until maybe next fall. But who knows. I played the Far Cry demo and had no problems at 1600X1200. If that is all they can throw at us for this year then maybe just maybe Ill have a computer that will last 24 months"

really? playing farcry demo no problems with 5900 at 1600x1200 res?
my friends and I tried that game, some had 9500 pro, 9700 pro, 9800p,5700fx u, 5900xt, 5900 ultra. myself I had a new 5900xt. with all the fx cards, we were all getting low scores using 4aa/8af with 1024x768 res at stock speeds. while all the ati cards performance were doing ok to really good in that game .

Even a 9500 pro was a bit faster than my 5900xt with those settings maybe it's because his system was a bit better than mine. at 1600x1200 without aa/af the 5900 ultra was getting average of 20 fps, and my friend has a better pc than yours. So I'm guessing either you turn all details down or you think running that game on your pc at 15-20 fps average is no problems.

now that a few of my friends and I, returned our nvidia fx cards and got ati's 9800 pro, were all so much happier playing most new and latest games with 4aa and 8af, than playing with our old fx cards, not to mention ati's image quality is better in 2d, it's more crisp, vibrant and clearer than nvidia's! I also noticed the fans on ati's are much quieter too.

 

[CaiNaM]

is there a way to accurately benchmark the farcry demo?

 

[CaiNaM]

Originally posted by: TheSnowman
i just went there because the results still hold true to more recent testing and wanted to copy/paste the results with the % difference tallys.

here is some recent shadermark tests.

thanks for the link, tho unfrotunately they are synthetic benchmarks. i did however find something worth noting further into the article. i found this comment regarding FarCry, which they descripbe as "a bright example of what we can expect from Hi-End boards with support for DirectX 9.0" (and something many of you ati fans have been touting as an example of ati's supreriority in ps2 shaders, which isn't really being contested.. but anyways..) interesting:

On the other hand,in the most playable resolution 1024x768x32 NVIDIA and ATI boards go abreast. And Gigabyte GV-N595U-GT even leaps forward thus justifying the name of the fastest board built on the NVIDIA GeForce FX 5950 Ultra chip.

so in other words, unless you want to accept the 30fps and below framerate at 1280x1024 and up, the 9800xt and 5950 run neck and neck around 40fps @ 1024x768.

kinda proves my point, thank you. using FarCry as an example, the ati is no more "futureproof" than nvidia - unless you condsider <30fps acceptable when labelling something as adequate hardware. 🙂

 

[kylebisme]

you are leaving out the fact that far cry forces nvidia to ps1.1 with the far cry demo though since ps2.0 tanks peformace on it, where with ati you get all the eye candy and same framerate as nvidia's ps1.1.

 

[CaiNaM]

Originally posted by: TheSnowman
you are leaving out the fact that far cry forces nvidia to ps1.1 with the far cry demo though since ps2.0 tanks peformace on it, where with ati you get all the eye candy and same framerate as nvidia's ps1.1.

well, i won't argue that, however the article does not specify one way or the other, in which case if indeed there was a "realworld" difference, those boys have no clue as to what they are printing, making everything in the article questionable - not just the parts that do not support your argument.

however, that's irrelevant to this argument anyways, as they make no mention of sub-par IQ on the part of the nv, making it just as playable on one platform as the other, still showing neither is more "futureproof" than the other, at least with realworld applications available today, and unless you are willing to accept sub 30fps frame rates.

 

[CaiNaM]

incedently, regarless of which shader version was used in the tests, this also supports rollo's stance that by the time dx9 apps are the rule, not the exception, neither of these cards will provide adequate performance (after all, the discussion wasn't even about the 5950u or 9800xt, and if the 9800xt barely offers playable framerates, where does the 9800pro stand?), and that "futureproof" is nothing more than hype or at best a term people use to make them feel better about their hardware.

 

[kylebisme]

Originally posted by: CaiNaM

well, i won't argue that, however the article does not specify one way or the other, in which case if indeed there was a "realworld" difference, those boys have no clue as to what they are printing, making everything in the article questionable - not just the parts that do not support your argument.

i just found the site googleing for shadermark, 5950 and 9800 when you complained about the date and circumstances of the previous numbers i posted. but ya they obviously aren't on the ball, those shadermark scores obviously are run at full precision when most people are reasonable enough to turn on partial precsion for the fx cards just so they don't look so incredibly bad at ps2.0. regardless, i don't see any reason to dismiss their benchmarks just because they arn't in the know on all the details. most people lack some knowledge about what they review in whatever area, you just have to know how to read the reviews for what they are worth.

Originally posted by: CaiNaM
however, that's irrelevant to this argument anyways, as they make no mention of sub-par IQ on the part of the nv, making it just as playable on one platform as the other, still showing neither is more "futureproof" than the other, at least with realworld applications available today, and unless you are willing to accept sub 30fps frame rates.

you can see image quality differences in the many shots that have been posted accross the net and here in this thread, you can even fire up the game and trying for yourself on both your cards. you can also force ps2.0 on your fx and see it tank to sub 30fps frame rates while your radeon keeps up much better.

Originally posted by: CaiNaM
incedently, regarless of which shader version was used in the tests, this also supports rollo's stance that by the time dx9 apps are the rule, not the exception, neither of these cards will provide adequate performance (after all, the discussion wasn't even about the 5950u or 9800xt, and if the 9800xt barely offers playable framerates, where does the 9800pro stand?), and that "futureproof" is nothing more than hype or at best a term people use to make them feel better about their hardware.

not at all, more future proof vs less future proof is something conservative shoppers keep in mind because it benefits them in the long run. like i just sprung for the audigy2 zx over the standard audigy2, not because i need 7.1 support as i have a 5.1 speaker setup and most games just use 5.1 anyway, but the zx may come in handy for me later down the road and even if it doesn't it will hold its resale value better. or back when i got a geforce3 instead of the geforce2 ultra that won quite a few benches at the time and cost a bit less to boot, sure the geforce2 ultra might have seemed like a good deal to a short sighted person; but in the long run the gefoce3 payed off as it was more future proof.

 

[CaiNaM]

well, i do have both cards, and i know their advantages and disadvantages, in both image quality and performance. i know how both run the farcry and ut demo, as well as daoc (which i am most familiar with, as it takes most of my gaming time), halo, call of duty, and bg1942 (desert combat).

there are some subtle performance differences in both performance and iq, however they are not all in ati's favor (but that's another discussion entirely). i overclock both cards, and my 5900 is FAR more overclockable than my 9800pro, which does indeed help make up for some of the 5900's defencencies.

as for your gf2/gf3 comaparison, that holds no relevance as they were different generations/architecure; not like the r3xx supports dx10 ;p

bottom line, and the point i've been arguing the entire time, is that while the ati is indeed the more capable part in dx9 (and nv has closed the gap somewhat, and may further with custom codepaths/optimizations), by the time dx9 titles form the majority of the market, neither card will be a very viable alternative. if the 9800xt, the fastest card available to us at this time, barely plays farcry at 40fps (and this is without aa/af remember), i just cannot see a vaild reason to think it will be viable in another 12-18 months when dx9 titles overtake dx7/8 titles. 2 years is LONG time for a gfx card these days, and the 9800p, while a great card, is already a year old.

in all the arguments you've made, i have seen nothing that would make me think otherwise, based on comparing both of these cards in everyday use. there is no reason for me to believe my 9800pro will stay in my pc any longer than my 5900. there is not "futureproof" when the industry is running 6-12 month product cycles, especially when applications requiring their hardware "features" are 1-2 years behind the hardware itself. put on your rose-colored glasses and believe your current generation is more "futureproof" than the next guys.. but you're simply kidding yourself.

 

[kylebisme]

some things are more futureproof than others, if you can't accept that then you are the one kidding yourself.

Originally posted by: CaiNaM
as for your gf2/gf3 comaparison, that holds no relevance as they were different generations/architecure; not like the r3xx supports dx10 ;p

but it does support dx9 were the fx cards have yet to get drivers to even fully acomplish that. furthermore, developers are going to be making games that run well on the next generation of midrage dx9 parts, and those parts will likely be about the same speed as the current high end from ati, if not slightly slower. that still leaves the current high end fx cards lacking, and the mid and low end ones in the dust.

 

[CaiNaM]

Originally posted by: TheSnowman
some things are more futureproof than others, if you can't accept that then you are the one kidding yourself.

Originally posted by: CaiNaM
as for your gf2/gf3 comaparison, that holds no relevance as they were different generations/architecure; not like the r3xx supports dx10 ;p

but it does support dx9 were the fx cards have yet to get drivers to even fully acomplish that. furthermore, developers are going to be making games that run well on the next generation of midrage dx9 parts, and those parts will likely be about the same speed as the current high end from ati, if not slightly slower. that still leaves the current high end fx cards lacking, and the mid and low end ones in the dust.

ahh.. but we haven't seen a huge difference in ps1.1 vs ps2; at least not yet. while it's "possible" your scenario may come true, there's nothing here now to supporting it, and again i point to the fastest dx9 card available today, the 9800xt, barely supporting playable framerates in FarCry, without the use of aa/af, which will further degrade performance, tho admittedly it will not impact performance as much as the nv parts.

and i do appreciate your arguments snowman; while you haven't shown me anything that would convince me rethink my position, i am learning something 🙂

 

[kylebisme]

Originally posted by: CaiNaM

ahh.. but we haven't seen a huge difference in ps1.1 vs ps2; at least not yet.

there is a huge difference between what can be accomplished with ps2.0 vs ps1.1, if you havn't seen it that is becasue you are not looking. if you are simply arguing that the difference isn't seen in many games that are out now, then again that not relvent to the topic at hand.

Originally posted by: CaiNaM
while it's "possible" your scenario may come true, there's nothing here now to supporting it, and again i point to the fastest dx9 card available today, the 9800xt, barely supporting playable framerates in FarCry,

your opinion of playable, while i imagine is similar to my opinion of playable, is not the majority opinion of playable. look at all the people playing ut2004 when that game drops into at least the 20s regardless of how fast a system it runs on. however, i get plenty playble framerate in far cry; i turn some options down to do it but the game still looks great, especialy some of the ps2.0 effects.

Originally posted by: CaiNaM
without the use of aa/af, which will further degrade performance, tho admittedly it will not impact performance as much as the nv parts.

actually aa and af shouldn't degrade peformace much on either line of cards, at least it doesn't on my rig. it seems that in many ways far cry is rather cpu bound, and shader bound in others, but that leaves plenty of room for aa and af.

Originally posted by: CaiNaM
and i do appreciate your arguments snowman; while you haven't shown me anything that would convince me rethink my position, i am learning something 🙂

it always feels good to be apprecated. 😀

 

[nRollo]

some things are more futureproof than others, if you can't accept that then you are the one kidding yourself.

The point is that while this is true, it's irrelevant for some of us. If the 5950 is at 20 fps average using PS2 in Far Cry, and the 9800 XT is at 40 fps, I wouldn't play in DX9 on either card as those framerates at those settings, are an unacceptable gaming experience.

10X7 with no AA/AF? The question becomes "what's more realistic- jagged, fuzzy, nicer effects- or DX8 with AA/AF for clear, smooth, duller effects?"

I don't know about you Snowman, but I see a big trade off here, and evidence the current generation of cards isn't equipped to handle DX9 PS2. If you have to run low res, give up AA, give up AF, I don't think you can honestly say one card is more "future proof". You're taking one step forward, and three steps back just to see that shinier pipe and shinier water.

 

[Gagabiji]

Originally posted by: Manzelle

Originally posted by: videoclone
Cindy22 Did you read the posts about NV40 having 16 Pipelines .... its looking like nvidia is cheating again and not playing fair because ATI's R420 only use's 8 Pipelines ... 😛 hehe lol

Ohh back to your flame bate forum topic whats more future proof ATI9800 or FX5900 well... both suck and in about 2 months they will be what the Geforce4 and ATI8500 are today !!! Crap ! .. but that?s how the computer industry has always been like nothings ever future proof !

I wouldn't call them crap...the 9700 Pro has been out for a long time and is still a very attractive card...

No-duh the 9700 Pro has been out for what now, 18 months? And it still is a VERY strong card.

 

[Genx87]

really? playing farcry demo no problems with 5900 at 1600x1200 res?

Yes np. I set everything to high except shadows and have AA in game set at medium. A little choppy when there is a lot of fighting on the screen for a few seconds but then it smooths out. As for fps I cant tell you what my actual fps are but is is smooth to the eye.

my friends and I tried that game, some had 9500 pro, 9700 pro, 9800p,5700fx u, 5900xt, 5900 ultra. myself I had a new 5900xt. with all the fx cards, we were all getting low scores using 4aa/8af with 1024x768 res at stock speeds. while all the ati cards performance were doing ok to really good in that game .

Well dont run with AA\AF enabled via the control panel. The demo makers have said it is borked in the demo per Digital life's review.

Even a 9500 pro was a bit faster than my 5900xt with those settings maybe it's because his system was a bit better than mine. at 1600x1200 without aa/af the 5900 ultra was getting average of 20 fps, and my friend has a better pc than yours. So I'm guessing either you turn all details down or you think running that game on your pc at 15-20 fps average is no problems.

I cant tell you what my fps are but the frames are smooth. Whether it is right at 24-30 it doesnt really matter as the thing is smooth to the eye. Except when there is heavy fighting then it lags for a few seconds. Also when you look at the wake on the boats.

now that a few of my friends and I, returned our nvidia fx cards and got ati's 9800 pro, were all so much happier playing most new and latest games with 4aa and 8af, than playing with our old fx cards, not to mention ati's image quality is better in 2d, it's more crisp, vibrant and clearer than nvidia's! I also noticed the fans on ati's are much quieter too.


Good for you. Apparently you and your friends must run a review site or something. I have never seen so many people have every single variety of FX cards and R3.xx cards to test with.
And not a single one could get far cry to run smoothly. Hmmmmm 🙂

And the whole 2d crisp argument is old news. People must have really bad monitors or something. I have a 9600 Pro AIW in my workstation and the 2d is no better than the GF2 MX that it replaced.

And now you are bragging about fans? I take it all of those 5900s you and your friends had were broken or something. I cant even hear the fan on my 5900 🙂

 

[rbV5]

Colin McRae Rally 04 performance test

I've been playing this the last couple days. The demo is a lot of fun, and plays well on my rig.

 

[nRollo]

re:Colin MaCrae


Of course, where they do note

With AA enabled, ATi's superior antialiasing implementation comes into play. Although in general the image quality is pretty close in the screenshot provided, there are areas where the ATi boards image is noticably better - Pay particular attention to the area directly underneath the spoiler, and the bottom of the roof section of the marquee.

they don't bother to re-iterate

The one disappointing thing about the graphics engine is the use of alpha textures in the trees and grass - As we know, these aren't antialiased by the multisampling method used by both ATi and nVidia, which leads to some nasty jaggies in places even with AA turned on.

that they noted earlier, and would not be an issue using nVidias non-multisampling AA modes.

 

[Acanthus]

My point with the 4 drivers comment was that i want to buy the game, put in the CD, install it, and play it.

I dont want to buy the game, put in the cd, install it, play it, have it crash, surf forums for hours figuring out whats wrong, then wait weeks/months for a new driver so i can play the game i bought.

 

[Hanners]

Originally posted by: Rollo
re:Colin MaCrae

they don't bother to re-iterate

The one disappointing thing about the graphics engine is the use of alpha textures in the trees and grass - As we know, these aren't antialiased by the multisampling method used by both ATi and nVidia, which leads to some nasty jaggies in places even with AA turned on.

that they noted earlier, and would not be an issue using nVidias non-multisampling AA modes.

Firstly, only MSAA is selectable from within the game itself, you can't specify whether SSAA or MSAA is used.

Secondly, performance would have been so poor using SSAA as to be worthless. It isn't like the 5900 had horsepower to spare once AA was enabled.

 

[rbV5]

Originally posted by: Hanners

Originally posted by: Rollo
re:Colin MaCrae

they don't bother to re-iterate

The one disappointing thing about the graphics engine is the use of alpha textures in the trees and grass - As we know, these aren't antialiased by the multisampling method used by both ATi and nVidia, which leads to some nasty jaggies in places even with AA turned on.

that they noted earlier, and would not be an issue using nVidias non-multisampling AA modes.

Firstly, only MSAA is selectable from within the game itself, you can't specify whether SSAA or MSAA is used.

Secondly, performance would have been so poor using SSAA as to be worthless. It isn't like the 5900 had horsepower to spare once AA was enabled.

Hey Hanners, you wouldn't have a comprehensive list of current and/or upcoming DX9 titles using pixel shaders would you?

The simple fact is - This game is fun to play.

Totally agree! I'll be picking this one up.

 

[CaiNaM]

Originally posted by: Rollo

some things are more futureproof than others, if you can't accept that then you are the one kidding yourself.

The point is that while this is true, it's irrelevant for some of us. If the 5950 is at 20 fps average using PS2 in Far Cry, and the 9800 XT is at 40 fps, I wouldn't play in DX9 on either card as those framerates at those settings, are an unacceptable gaming experience.

10X7 with no AA/AF? The question becomes "what's more realistic- jagged, fuzzy, nicer effects- or DX8 with AA/AF for clear, smooth, duller effects?"

I don't know about you Snowman, but I see a big trade off here, and evidence the current generation of cards isn't equipped to handle DX9 PS2. If you have to run low res, give up AA, give up AF, I don't think you can honestly say one card is more "future proof". You're taking one step forward, and three steps back just to see that shinier pipe and shinier water.

i still maintian there is no such thing as "future proof".. the future always comes, and the past is always left behind.

if i had to concede that there was acutally such a thing between these 2 cards, i'd have to maintain the degree of "future proof" is more deteremined by how low one's standards are than the hardware itself. 😛

 

[kylebisme]

but then it is just as easy to turn that around and say that the unwillingness to accept the concept of futureproof as a relative term is a blatant sign of one's inability to reach a state of emotional fulfillment. 😉

 

[BFG10K]

we can use more hd and ram TODAY with significant benefits.

If he was using it then it wouldn't be reported as spare. Unless your're saying it's good just in case I need it, in which case exactly the same thing applies to PS2.0.

further, defining the significance that today's performance will hold in future, upcoming times and predicting how long these current parts will adequately play those "future" games is even more of a stretch.

And how fast will his rig handle games that fill up his entire RAM and HD space? Not very well I'd imagine, so using your logic that storage space is not needed and therefore it should be removed.