GF FX 5800 Ultra Vs. Radeon 9700 Pro

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
Originally posted by: Rollo
Rampaging Rogo:
"Reading THIS review I would have to conclude that the FX is not holding up so well."
You can't just capitalize a word to make it work as a link.

"Actually it was always compared to the geforce 4 ti series, even though it was aimed at the Geforce 3 Ti500."
Bzzzt. Wrong. The First R8500 review on this board, 10/17/01 was out long before the First Ti review,2/6/02, four months later. So there's four months they weren't compared to Tis. After that they were for a while, only because ATI had no comparable part to the Ti generation. (much the same as the last half year where the Ti4600 has been compared to the 9700Pro)

"So chizzow, don't think you have a monoply on knowlege alright buddy"
He's right about this though. The Ti4200 came out months after the 4400/4600, and no one really considered 8500s competition for them.

"Reviews of the 8500 were springing up in late december and reviews of the Geforce 4 were out in january of 2002"\
Hmmm. Not on this board, see above.

"In those benchmarks the radeon is faster than the geforce 3 ti500 and relatively close to the geforce 4 ti4600.
So what is your point again big guy?"
Tonite my point is you must have put Quaaludes in your coffee. An R8500 close to a Ti4600?!
20%slower at 12X10, 30% slower at 16X12
55% slower at 10X7, downhill from there
25% slower and worse at hi res
32% slower here

If the 9700/9800 could pull off victories like these at anything except high res, very high AA/AF, I'd believe you guys.

"I had personall exprience with two geforce 4 cards and a radeon 8500LE
My own judgment (and perception) is more valid than anybody's post on a public form (to myself) or an online review.
The fact that I kept the 8500 and sold both the geforces is a testament to the 8500s staying power (only to me).
If you have run both the cards then I would be more than happy to hear what you have to say about them. As it stands I don't have any grounds to give a flying fook what you say."

Uh oh, careful what you wish for. I have owned and used a Ti4200, Ti4400, and a R8500 retail. The R8500 retail only lasted 3 weeks, because I got tired of it locking up in games after fdisk, fresh install, many drivers, bios flashes, etc.. I popped in the the Ti4400, did none of the above, and it ran FLAWLESSLY. (as did the GF3 I had before the R8500)
Willikers! Same box, and I had problems with my ATI, but not nVidia. In those days, the big workaround on Rage 3d was underclock your retail to LE speeds, and sure enough, that did seem to lessen the problem. (and performance)

why are you linking to Q3? a game that at 100fps is fine and who gives a crap about 200fps vs 220fps?

Also if you look at the various reviews around they show more than just high rez and AA/AF situations where the 9700Pro is beating the FX. If you even bother to follow the links posted you can crearly see the truth. I trust anandtech, hardocp more than Toms hardware reviews and definately more than your word which you seem to take as gospel for everyone.
 

Rogozhin

Senior member
Mar 25, 2003
483
0
0
Rollo

"Uh oh, careful what you wish for. I have owned and used a Ti4200, Ti4400, and a R8500 retail. The R8500 retail only lasted 3 weeks, because I got tired of it locking up in games after fdisk, fresh install, many drivers, bios flashes, etc.. I popped in the the Ti4400, did none of the above, and it ran FLAWLESSLY. (as did the GF3 I had before the R8500"

Why be such a pompus azz? I don't go around spounting my crappy exprerience with MY gainward geforce 4 and msi geforce 4 as the be all and end all of geforce 4 reliability and performance, so why do you?

I know you're older but that does not make you the "prime mover."


And I didn't wish for a gawd dam thing.

What would make you say such a thing?

Rogo
 

BFG10K

Lifer
Aug 14, 2000
22,709
3,007
126
You would run "balanced" on the FX BFG; and compare it to your "quality".
Uh no, I wouldn't because if you don't use the application setting the driver starts to cut corners.

Not everyone is as convinced as you are that ATIs "performance settings (actually any settings) for AF don't result in image degradation.
Performance is not as good as quality but the difference is tiny to non-existant in practice.

Digit Life shows how "accurate" our 9700's AF really is- and not with a screenshot of Quake
Instead of relying on one screenshot from a random site why don't you test it for yourself like I've done? I've tested performance vs quality in over forty games and I have effectively found zero difference in practical terms, except of course the inferior performance that quality has compared to performance.

Also news flash Rollo: the NV30's anisotropic isn't 100% accurate either. Of course it doesn't seem to make it any faster either so you really have to wonder exactly how nVidia could botch it up so badly.

Again, "balanced" is nVidia's "quality" setting,
No it isn't Rollo, that's total bullsh*t. Application is the quality setting; any other setting reduces image quality by cutting corners in the rendering.

What a joke. A fps player would NEVER use those settings, because even mediocre players like me would have plenty of time to dance around him blasting away as his overloaded VGA stuttered.
212 FPS average in Quake 3's demo four @ 1600 x 1200 x 32 with 16x performance anisotropic?
136 FPS average in Quake 3's demo four @ 1792 x 1344 x 32 with 16x performance anisotropic?

No, the only joke here is you using your 1280 settings on a $400 card because you'll scared you'll break it or something.

No I don't buy high end VGAs to run games all choppy and jerky like you apparently do.
Uh no, I do no such thing. You must have something really screwed with your system and/or you have no idea how to setup your video card. In either case it appears to have a skewed and unreleastic view of the card and you're using that view to attempt to place the FX on a pedestal that it doesn't deserve to be on.

The 9800 is probably the best card all around, the 5800 Ultra second, our 9700 Pros lingering in 3rd.
You know, I'm having a tough time figuring out whether you're just trolling to wind up the forums here or whether you actually believe that what you're saying is true.

To spare yourself further embarrasment I suggest you look at Oldfart's results to see just how badly your "second best" card gets beaten, even at the low resolution 1024/1280 results that you seem to enjoy using. And yes, the only valid ones are the application settings for the NV30; there's a reason why the other scores are faster you know.
 

BFG10K

Lifer
Aug 14, 2000
22,709
3,007
126
Tonite my point is you must have put Quaaludes in your coffee. An R8500 close to a Ti4600?!
Yeah, in some cases the 8500 is actually faster than the Ti4600 and in most cases it's equal to or faster than a Ti4200. Pay attention to the 1600 x 1200 benchmarks, especially the ones with minimum scores.

Driver advances from ATi have made the 8500 an approximate equal to a Ti4200, losing in fillrate limited situations but winning, where it counts, in memory bandwidth limited situations. That is unless you're Rollo who enjoys gaming at VGA resolutions on $400 video cards.
 

nRollo

Banned
Jan 11, 2002
10,460
0
0
"Yeah, in some cases the 8500 is actually faster than the Ti4600 and in most cases it's equal to or faster than a Ti4200. Pay attention to the 1600 x 1200 benchmarks, especially the ones with minimum scores."
LOL- one benchmark on one site and the 8500 is the equal of the Ti4600. Sure BFG, and how many cups of the 'lude coffee did you have?
Shall I post links to a couple hundred benchmarks showing the 4600 beating the 8500, or would that not be sufficient counter point to your one?
rolleye.gif


You'll post just about ANYTHING, no matter how ridiculous and far fetched, just to disagree with me. Probably because you're tired of me pointing out the errors in your "serious" posts, and how much you exaggerate and use specific criteria others may not deem relevant to make your points.
 

nRollo

Banned
Jan 11, 2002
10,460
0
0
Wow. I can believe you own a coffee house, drink a bit of caffeine Rogo, you are one prolific poster. I'll try to address some of your many points:

"You buy an fx for these reasons only.
1.You want to fiddle with new tech (i would myself, short on money at the moment)
2.You are truely a nvidiot and don't care about price.
3.You simply don't know any better."
Disagree. You might buy one because you don't like the rolling gray bars, or ATI incompatibilities/driver issues.

"Actually it was always compared to the geforce 4 ti series, even though it was aimed at the Geforce 3 Ti500"
That would be a good trick as the 8500 was out before the GF4. How did they compare when no one had a GF4? Speculation?

"Did you look at my link???? "
No. It doesn't show up as a link on my Win ME IE , all the others seem to. Yours looks like a capitalized word.

"I had an LE and after a bios flash and volt mod I was running 300/300 with no artifacts no bsods and no lock ups"
OK, I did though, as did 100s of others on Rage3d. That card was flakey and not worth the trouble.

"I remeber in one of your posts that you don't believe that anand is a very good source of benchmark info. I will go find it. So why quote him"
I didn't say he's bad, I said there are things to consider other than what he reviewed for the 5800 Ultra.

"My point is that the 8500 WAS benched against the Geforce 4 series as soon as they came out. "
Like I said, ATI had nothing else to bench. Just like all the reviews comparing the Ti 4600 to the 9700.






 

Rogozhin

Senior member
Mar 25, 2003
483
0
0
"Actually it was always compared to the geforce 4 ti series, even though it was aimed at the Geforce 3 Ti500"

Sorry about this one, I was typing like crazy, I should have said "as soon as the geforce 4 series was released"


Rogo
 

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
Originally posted by: Rollo
Wow. I can believe you own a coffee house, drink a bit of caffeine Rogo, you are one prolific poster. I'll try to address some of your many points:

"You buy an fx for these reasons only.
1.You want to fiddle with new tech (i would myself, short on money at the moment)
2.You are truely a nvidiot and don't care about price.
3.You simply don't know any better."
Disagree. You might buy one because you don't like the rolling gray bars, or ATI incompatibilities/driver issues.


er...who has those problems? I've built 5 systems using 9700Pro and NOT ONE has that issue. Did you even try the cat3.2 driver? Or are you installing from the Cd and saying..."it sucks" and uninstalling?
 

nRollo

Banned
Jan 11, 2002
10,460
0
0
CMDREDD:
"why are you linking to Q3? a game that at 100fps is fine and who gives a crap about 200fps vs 220fps?
Also if you look at the various reviews around they show more than just high rez and AA/AF situations where the 9700Pro is beating the FX"

Better question: Why are you missing the whole point of the post you quoted. We're talking about old R8500 reviews here CM. Back in 2001, Q3 benching was all the rage....

"er...who has those problems?"
LOL
93 threads, 100s of posts about the lines

"I've built 5 systems using 9700Pro and NOT ONE has that issue. Did you even try the cat3.2 driver? Or are you installing from the Cd and saying..."it sucks" and uninstalling?""
Oh, I did all kinds of things CM. I've been building my own computers for gaming since the early 90s, built/repaired many boxes, and have made my living in IT/IS the last five years. I figured out how to install a video card long ago.
rolleye.gif

Here's the neat part CM:
On the exact same box, I got the rolling lines on 2/3 monitors I own! Jinkies, lucky for me I had three to try, eh?

 

bunnyfubbles

Lifer
Sep 3, 2001
12,248
3
0
Is it just me or does Rollo come across as believe everyone with a 9700 has problems with rolling bars? Obviously the majority doesn?t otherwise the 9700 would be a flop. Usually if you have a problem you'll post about it. If everyone who owned a 9700 and didn't get rolling bars and bothered to make a post about it, I'm sure it'd be nearly impossible to weed out the threads complaining about the problem from the scores of threads stating that they did not have a problem.

Using the rolling bars against ATI is almost as bad as continuing to put them down for poor drivers/driver support. Oh but wait you do that too. Too bad ATI has really picked up the pace and their drivers and support has really improved over the past several months, and to think that the rolling bars does not affect an insane amount of users and that the issue is most likely being worked on to solve...

Here's the neat part CM:
On the exact same box, I got the rolling lines on 2/3 monitors I own! Jinkies, lucky for me I had three to try, eh?

You must not be have a logical knack for sciences and statistics considering there are hundreds of possible variables in a computer that could cause or support the problem of rolling bars. Changing monitors won?t be a guaranteed success when the problem could lie the card, motherboard, power supply to the card/monitor/motherboard/AGP slot... Heck, the problem could just be your personal location and interferences. Oh and you keep talking about how you bought a 9700 a while back for $380 or whatever. Guess that makes you bought the card early on in its life, certainly not a more mature one which would most likely be bound to have fewer problems. Your rolling bar problem very well could be primarily due to inadequacies of your relatively immature card.
 

BenSkywalker

Diamond Member
Oct 9, 1999
9,140
67
91
Too bad ATI has really picked up the pace and their drivers and support has really improved over the past several months

Like the Cat3.2s that only broke ~half a dozen games?

Half-Life- Hard Lock
CounterStrike- Hard Lock
BlueShift- Hard Lock
TFC- Hard Lock
Opposing Force- Hard Lock
Sac- Serious Image Corruption(actually, this one has been around since the launch of the R300, not a new bug)

Then you have the performance on the older Lithtech games. NOLF ran better on my GF2 then R9500Pro, KPC was unplayable using the R300(stuttering so badly it appeared to be running ~@10-15FPS running 800x600x16 w/o AA or AF). Quake3 and JKII show image corruption with pixel popping along the seems of large polys. I couldn't ever get the S3TC textures to work in UT either, although I gave up fairly quickly when I realized how 'great' ATi's drivers were so this one may have been something I did. I traded off my R9500Pro mainly because of ATi's wonderful drivers. If every game their boards ran could be played @1600x1200x32x6x16 300FPS it wouldn't matter for all the games that won't play at all or have varrying degrees of image corruption.
 

kylebisme

Diamond Member
Mar 25, 2000
9,396
0
0
Originally posted by: BenSkywalker
Like the Cat3.2s that only broke ~half a dozen games?

Half-Life- Hard Lock
CounterStrike- Hard Lock
BlueShift- Hard Lock
TFC- Hard Lock
Opposing Force- Hard Lock

wow, wat to nail them 5 times for one mistake there Ben! were you on the committee that claimed Saddam posed an immediate and serious threat with stockpiles of WMD, or did they just have someone else with your talents? :D
 

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
Originally posted by: Excelsior
You do know 5 of those use the HL engine and/or are mods for HL.

and Ut2003, Q3, RTCW, Medal of Honor, BF1942 all run fine. and these games are actually worth playing on new hardware cause they look good.
 

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
Originally posted by: BenSkywalker
Too bad ATI has really picked up the pace and their drivers and support has really improved over the past several months

Like the Cat3.2s that only broke ~half a dozen games?

Half-Life- Hard Lock
CounterStrike- Hard Lock
BlueShift- Hard Lock
TFC- Hard Lock
Opposing Force- Hard Lock
Sac- Serious Image Corruption(actually, this one has been around since the launch of the R300, not a new bug)

Then you have the performance on the older Lithtech games. NOLF ran better on my GF2 then R9500Pro, KPC was unplayable using the R300(stuttering so badly it appeared to be running ~@10-15FPS running 800x600x16 w/o AA or AF). Quake3 and JKII show image corruption with pixel popping along the seems of large polys. I couldn't ever get the S3TC textures to work in UT either, although I gave up fairly quickly when I realized how 'great' ATi's drivers were so this one may have been something I did. I traded off my R9500Pro mainly because of ATi's wonderful drivers. If every game their boards ran could be played @1600x1200x32x6x16 300FPS it wouldn't matter for all the games that won't play at all or have varrying degrees of image corruption.

for starters I don't have problems and why have none of the reviews mentioned problems in Q3 and JK2? and in UT the S3TC textures only work in OGL and the OGL.dll didn't work quite like it should with the ATI cards. I did get it to work after using a custom dll file
 

nRollo

Banned
Jan 11, 2002
10,460
0
0
"You must not be have a logical knack for sciences and statistics considering there are hundreds of possible variables in a computer that could cause or support the problem of rolling bars. Changing monitors won?t be a guaranteed success when the problem could lie the card, motherboard, power supply to the card/monitor/motherboard/AGP slot... Heck, the problem could just be your personal location and interferences. "
You know what's funny about that Bunny? I ran a bunch of other cards in THAT EXACT SAME COMPUTER WITHOUT ANY ROLLING LINES.
As far as my "knack for the sciences" goes, familiar with the term "ceteris paribus"? Sometimes "more scientific types" use it in their experiments, I've heard.
rolleye.gif

Anyway- here's the deal "scientist":
I put the R9700Pro in it. Rolling lines. I try the DVI adapter trick, which fixes it for the most part. But I'm not satisfied! So- I fdisk/format/clean install- lines. I say WTF? This same computer, on clean installs, has run a R7500, a R8500, a GF3, a GF4Ti, and a V5 without lines. I pop the 9700Pro out, Ti4400 in, no lines, I put the R9700 Pro back in lines. I put in a GF3, no lines. I put the R9700 Pro back in -lines.

It must be my power supply. Good lord. Yet again you prove why the saying isn't "smart bunny".
rolleye.gif


BTW Bunny- my stuff is all a pretty good stuff: Antec 430 True Power (you spend $100 on your psu Bunny?) Asus P4Pe deluxe, P4 2.53, 512 Crucial PC2700, etc.. I can see why you'd theorize that stuff is at the root of all seemingly ATI related issues...
rolleye.gif

 

nRollo

Banned
Jan 11, 2002
10,460
0
0
"Why aren't people comparing the 5800 to the 9800...it only seems fair"
It's not fair! The 9800 came out later!

LOL

I don't know, I compare all four personally. (5800/Ultra, 9700Pro/9800Pro) I think you're seeing more people debate the 9700Pro vs FX Ultra because lots of people have 9700Pros and want to compare them to FX Ultras?
 

oldfart

Lifer
Dec 2, 1999
10,207
0
0
Originally posted by: cmdrdredd
Originally posted by: BenSkywalker
Too bad ATI has really picked up the pace and their drivers and support has really improved over the past several months

Like the Cat3.2s that only broke ~half a dozen games?

Half-Life- Hard Lock
CounterStrike- Hard Lock
BlueShift- Hard Lock
TFC- Hard Lock
Opposing Force- Hard Lock
Sac- Serious Image Corruption(actually, this one has been around since the launch of the R300, not a new bug)

Then you have the performance on the older Lithtech games. NOLF ran better on my GF2 then R9500Pro, KPC was unplayable using the R300(stuttering so badly it appeared to be running ~@10-15FPS running 800x600x16 w/o AA or AF). Quake3 and JKII show image corruption with pixel popping along the seems of large polys. I couldn't ever get the S3TC textures to work in UT either, although I gave up fairly quickly when I realized how 'great' ATi's drivers were so this one may have been something I did. I traded off my R9500Pro mainly because of ATi's wonderful drivers. If every game their boards ran could be played @1600x1200x32x6x16 300FPS it wouldn't matter for all the games that won't play at all or have varrying degrees of image corruption.

for starters I don't have problems and why have none of the reviews mentioned problems in Q3 and JK2? and in UT the S3TC textures only work in OGL and the OGL.dll didn't work quite like it should with the ATI cards. I did get it to work after using a custom dll file
I dont have HL or any of its mods installed. After my last reformat, I never got around to putting it back in. I've moved on to newer games. Maybe I'll install HL (not all the mods) and see how it is. They are all the same engine. If one works/doesn't, they all ar in the same boat.

Same with NOLF. Haven't played it in ages, and dont have it installed. I guess I could give it a quick whirl as well

UT with S3TC works just fine

Q3, dont see any image problem. Looks great!

EDIT Tried out HL. I installed HL, the high def pak and patched it to 1.08. Played it for awhile (OH the fond memoriies!!) I just loved that game!! Anywho, it played just fine. Had it maxed out as possible with 1290 x 960 4X AA (I guess I could have used 8X) and 16X AF. There is a black screen problem when I exit the game. Didn't mess with it much to see how to get around it.

EDIT 2
Nolf also works fine. Frame rates are fast according to FRAPS. Running 1152 x 864 4X AA 16X AF, all game options set to max, my frame rate never dropped below 100 (my refresh rate) Vsync was on.




 

BFG10K

Lifer
Aug 14, 2000
22,709
3,007
126
LOL- one benchmark on one site and the 8500 is the equal of the Ti4600.
LMAO!

HINT: click on the "next" button at that website.

Shall I post links to a couple hundred benchmarks showing the 4600 beating the 8500,
You could but they'd be invalid in this discussion since they don't disprove that the 8500 can beat the Ti4600 at times, which is exactly what I was saying happened.

You'll post just about ANYTHING, no matter how ridiculous and far fetched, just to disagree with me.
Do you deny that the 8500 can beat the Ti4600 at certain times? If not then what are you doing here? If you are I'd like to see a logical and rational explanation that discounts the benchmarks I provided.

Like the Cat3.2s that only broke ~half a dozen games?
Please don't try to pretend that nVidia's drivers are perfect. Even the reviewers themselves have been commenting on performance issues and rendering errors and usually they don't spot things like at all. Also how's that WHQL certification coming on nVidia's drivers? Are we going to get it sometime this year? Or do we wait until 2004?

NOLF ran better on my GF2 then R9500Pro,
NOLF's performance is great though I'll admit I do have a laggy mouse. It's certainly an issue ATi needs to fix.

Quake3 and JKII show image corruption with pixel popping along the seems of large polys.
Never had any issues, even on the CD Catalyst 2.2s.

I couldn't ever get the S3TC textures to work in UT either,
Use the latest renderer from Epic (the one posted in October). It was some issue with parsing OpenGL strings I believe. Also I'm sure you enjoy using 16 bit S3TC textures on your Ti4200 while my Radeon looks awesome, especially in the sky.
 

BenSkywalker

Diamond Member
Oct 9, 1999
9,140
67
91
TheSnowman-

wow, wat to nail them 5 times for one mistake there Ben!

Say nVidia had one bug that caused all OpenGL games to cease working, I'm sure you would have the same attitude wouldn't you? They wrecked 'em all, they should be held accountable. The fact that one bug could break several of the most popular PC titles ever and to date and ATi didn't catch is speaks volumes about the QC of their driver team.

Oldfart-

There is a black screen problem when I exit the game. Didn't mess with it much to see how to get around it.

Odd, it is a known and admitted bug(by ATi) that the game hard locks systems whenever you try to go from the game to the menu running OGL. Only thing yet to manage to hard lock my new build, ATi's drivers. Surprised you don't have them installed, next months announcement got me wanting to play through them all again :)

Nolf also works fine. Frame rates are fast according to FRAPS. Running 1152 x 864 4X AA 16X AF, all game options set to max, my frame rate never dropped below 100 (my refresh rate) Vsync was on.

Try turning around quickly ;) BFG has the same issue it seems. It doesn't matter what settings you use, from 640x480x16 to 16x12x32x6x16. NOLF2 does not have the issue(although KPC which also runs on the older Lithtech engine does).

Q3, dont see any image problem. Looks great!

It's pixel popping at the seems of large polys, there was a thread posted by another user with screenshots to go along with it. Happens in JKII also.

BFG-

Even the reviewers themselves have been commenting on performance issues and rendering errors and usually they don't spot things like at all.

Like the SeriousSam issue? I added two lines to my SS config file and that went away(game bug obviously). What are the other ones? The problems with the FX boards? Haven't run into them with the GF4.

Use the latest renderer from Epic (the one posted in October).

I tried. I have all of the renderers(even the one Gustafson wrote before Vogel took over), couldn't get it to work with any of them but as I said, I simply gave up. There were too many more serious issues that I had. UT did run just fine on the R300 without the high res textures, too many other games I had didn't.

Edit-

Forgot to add, I try and avoid WHQL drivers with nV boards as I always have. They do not include OCing utilities(as per WHQL requirements).

ZimZum-

I was commenting on ATi's driver team improving over the last few months. They released a driver that broke a lot of games what worked fine with the previous version, including the most popular online game in the world. Seems to me more like they have fallen quite a bit back in the last couple of months then really improved. I'm not saying nV is perfect, nor that they have really improved over the last few months, but all of my games now run.