• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Difference between 5900 nu, 5900 ultra and 9700 pro?

the difference between nvidia's 5900 fx and ati's almost 2 year old 9700 pro is that the 5900 handles directx 9 games poorly.

the 9700pro or 9800 pro can handle new and future diretx9 games much better than 5900fx specially when using aa and af with nvidia's forceware driver cheats disabled

 
you can also use trilinar filtering with a 9700pro. 🙂

but to be fair, the 9700pro isn't nearly 2 years old.
 
Having read this AnandTech article, then looking here in the Video forum at 5900 overclocking success (5900 being the Non-Ultra, 128MB card with 400MHz core/850MHz RAM), I went ahead and bought a plain-jane XFX GeForceFX 5900 and here's the overclock I got (I use the 53.03 Drivers) I realize my 5900 doesn't have the 256MB of memory that the 5950 Ultra referenced in the first link does, however, I think if the performance is comparable with even the 5900 Ultra's numbers I did just fine with my sub-$180 purchase 🙂

Peace
 
yeah but Nvidia has been trying to force out performance that the rendering gets worst... 🙁
so much for desperation to hog the poll position
 
In dx9:
9700Pro>5900U>5900

OpenGL (will have to see what happens in doomIII 😀):
5900U>5900>9700Pro

In DX8:
5900<9700pro<5900ultra

I'd get 9700pro for it's high dx performance, specially dx9. And it performs very decently in OpenGL to
 
I'd get 9700pro for it's high dx performance, specially dx9. And it performs very decently in OpenGL to
I'd get 5900 for it's high ogl perofmormance, specifically doom3. And it performs very decently in dx9 too
 
In DX8:
5900<9700pro<5900ultra
You have a link or two backing this up McArra?

UT2003 is a DX8 game, and the 5900 is beating the 9700Pro at every resolution Here

Splinter Cell is DX 8 as well I believe, and the 5900 > 9700Pro at every resolution againMore proof

IL2Sturmovik is DX8 as well, and, you guessed it, 5900 beats 9700 Pro at every resolution.
So much misinformation here

Errrr, Nascar 2003 is DX8, and the 5900 stomps a 9700Pro at the first two, wins the third, and only loses the 4th by 1fps.
McArra, what the heck are you talking about?

So I just linked a direct comparison of the 5900 and 9700Pro in four popular games, at four popular settings, and the 9700Pro won 1/16 benchmarks. (and that one by 1 fps! Woot!) I can see why you say 5900<9700Pro, the 9700Pro is faster a whopping 6% of the time!


Like I've said before, unless you're spending much more on a 9800Pro, the 5900 is probably the best "bang for buck" card right now.


 
Appreciate everyones thoughts...
Like I've said before, unless you're spending much more on a 9800Pro, the 5900 is probably the best "bang for buck" card right now.
And thats what I'm looking for...Not into spending $400 if I can help it...but want *great* performance..
 
I'm getting the Geforce 5900 for OpenGL peformance.

But I guess I would have to recommend to get the 9700 Pro because it offers better DX9 performance and great OpenGL performance.

nVidia drivers do not cheat Cindy, go research.
 
Originally posted by: Rollo
In DX8:
5900<9700pro<5900ultra
You have a link or two backing this up McArra?

UT2003 is a DX8 game, and the 5900 is beating the 9700Pro at every resolution Here

Splinter Cell is DX 8 as well I believe, and the 5900 > 9700Pro at every resolution againMore proof

IL2Sturmovik is DX8 as well, and, you guessed it, 5900 beats 9700 Pro at every resolution.
So much misinformation here

Errrr, Nascar 2003 is DX8, and the 5900 stomps a 9700Pro at the first two, wins the third, and only loses the 4th by 1fps.
McArra, what the heck are you talking about?

So I just linked a direct comparison of the 5900 and 9700Pro in four popular games, at four popular settings, and the 9700Pro won 1/16 benchmarks. (and that one by 1 fps! Woot!) I can see why you say 5900<9700Pro, the 9700Pro is faster a whopping 6% of the time!


Like I've said before, unless you're spending much more on a 9800Pro, the 5900 is probably the best "bang for buck" card right now.

Sorry I didn't thought a lot about it. Just remenbered GamePC's reviews where non-standard games are run and 9700pro/9800np take the lead in quite benches, but I don't seem to find it. It may was another page.
 
No, you're just right. The 9700Pro is faster than the 5900NU most of the time. 😛

benchmarks

Rollo, that review you posted is comparing 3 differnet 5900 Ultra cards to the 9700Pro! I don't know why they label it as an "FX5900-TD128", but look at the summary of which cards they're testing:

cards tested

Quite definitely 3 different 5900Ultras, a 5600Ultra, and a Ti4600 on the NVIDIA side up against a 9800Pro256MB, 9800Pro128MB, and 9700Pro. Of *course* the 5900Ultra beats the 9700Pro, but that's not quite what you were trying to prove.
 
Ya Matt I was just noticeing the same: Actually 4 ultras.
System Setup

Intel Pentium 4 3.2GHz

ASUS P4C800 Deluxe

512MB OCZ EL PC3200 (DDR400) SDRAM

ATI RADEON 9800 PRO ? 256MB
ATI RADEON 9800 PRO ? 128MB
ATI RADEON 9700 PRO
Driver version CATALYST 3.6

ASUS V9950 Ultra
eVGA e-GeForce FX 5600 Ultra
Leadtek A350 UltraTDH MyVIVO
MSI NBox N5900 Ultra
NVIDIA GeForce4 Ti 4600
Driver version Detonator 45.23


No nu here... Rollo?
 
I found this but is quite old. It is when 5900U arrived and 9700Pro gave it some hard competition (cat 3.5 times). Link

Another one (much newer from Matthias's link): Link
 
Yes, but one of the "Ultras" is a 5600 Ultra. 🙂

Here's their *actual* review of the MSI FX5900-TD128:

review

Note the 9700Pro beating it handily in UT2K3 *without* AA/AF (which is why the results are so close here). They must have had a typo in that other article, since this card is nowhere near as fast as a 9800Pro.

Edit:

Here's the stock 5900 losing to the stock 9800NP (about the speed of a 9700Pro) in UT2K3 with AA/AF on.

link

The 5900NU may have an edge in certain OpenGL titles, but the 9700Pro beats it pretty solidly in DX8 games with AA/AF on.
 
Originally posted by: Matthias99
No, you're just right. The 9700Pro is faster than the 5900NU most of the time. 😛

benchmarks

LOL, the 9700 Pro isn't listed in this benchmark. Either you gave us the wrong link or you're totally blind. 😛
 
Originally posted by: bpt8056
Originally posted by: Matthias99
No, you're just right. The 9700Pro is faster than the 5900NU most of the time. 😛

benchmarks

LOL, the 9700 Pro isn't listed in this benchmark. Either you gave us the wrong link or you're totally blind. 😛

9800 is, wich performs as 9700Pro.
 
*sigh* Thank you for the link, it showed that the 5900 beat out the 9800np (or 9700 Pro if you want to call it that) in five of the seven different programs they used in the article.
 
Rollo, that review you posted is comparing 3 differnet 5900 Ultra cards to the 9700Pro! I don't know why they label it as an "FX5900-TD128", but look at the summary of which cards they're testing:

LOL, I stand corrected on the benchmark issues. (and somewhat surprised they mis-labelled all the benchmarks!)

In any case, the 5900NU is under $200, and the 9700Pro is $238 apparently, so that is a 20% price difference for performance that is only noticeably worse in DX9, PS2.0 shaders. I don't consider ANY card running DX9 PS2 shaders acceptably these days and will probably run HL2 at PS 1.4 if I haven't upgraded my 9800Pro by then. On the other hand, Doom 3 should run much better on a 5900NU than a 9700 Pro, so there is that to consider.
So I'll stick with my 5900 NU recommendation in this case, especially since I seriously have to wonder if there is a 5900 NU on the planet that cannot make that huge leap from 800/850 to 850/850 and achieve full 5900U status. (and then my benchmarks would be valid, although you'd have to consider the 9700Pros OC potential as well)

BTW- Doom3 is far more interesting to me than HL2, so I'd lean nVidia on this one. If HL2 were more important to me, I'd go ATI.
I think all these cards (5900s/9800s) are close enough in performance you'd be hard pressed to tell the difference, except in pretty specific instances.
 
Originally posted by: Rollo

[snip]
especially since I seriously have to wonder if there is a 5900 NU on the planet that cannot make that huge leap from 800/850 to 850/850 and achieve full 5900U status. (and then my benchmarks would be valid, although you'd have to consider the 9700Pros OC potential as well)
[snip]

I can think of at least one. Not to mention that OC'ing voids your warrantee and could possibly put you out $200+ on the card - so it really shouldn't be used in a comparison.

Good thread BTW, lots of good information.


Kudos,
Chum
 
Nitpicks:

UT2K3 is DX7-level tech. SC was developed for the Xbox, so no doubt it's optimized for nV's GF architecture. So neither benchmark proves the FX's inherent superiority over ATi's DX9 cards, but I suppose most ppl don't care what's in the black box. The IL2 numbers you linked show the 9700P losing by only 10% to the 5900[U, apparently], and that's not factoring in ATi's superior AA quality. nV just seems to be doing better on Nascar2003 (though it's odd that the Radeons all stay below 60fps and lose performance much less at higher res, which may indicate a driver limitation).

I just linked a Digit-Life review in another thread that compared a Gainward 5900 (GS 1200) to a 5900U. Even when clocked to 5900U levels and beyond, the Gainward couldn't catch up to the 5900U, so it's not at all a given that just OC'ing a 5900 to 5900U speeds will result in 5900U performance. OTOH, I think that review was fairly old, so maybe newer drivers make better use of OC'ed speed.
 
Back
Top