Woah...

[NYHoustonman]

Huge drop in prices for the FX 5900 128 on pricewatch the past few days. It's apparently now down to $290 shipped retail box, from around $340 last time I checked. Anybody have any idea why, and if I should now consider this more seriously vs. the 9800 pro 128, which was what I was leaning towards because it was cheaper? Is it worth the extra $30 bucks? It was a done deal before this, but now I dunno. Is all this stuff about the 9800 being better with shadows going to affect performance in future games (mainly HL2) alot, or is it just a load of bs?

 

[Corporate Thug]

wow, thats kinda weird. too bad both places under $300 is in CA

i'm really impressed with the review of the evga 5900 ULTRA. i wonder if this can OC to ultra speeds

 

[JulesMaximus]

From what I've seen in benchmark testing there isn't a heck of a lot of difference between the FX 5900 and FX 5900 Ultra in performance. Someone correct me if I'm wrong?

 

[CVNet1]

I wouldn't put much faith in benchmarking (ie 3dMarking) anymore as it has been shown that driver "Optimizations" can easily make a products score increase by a significant amount. And even discounting that, the latest Futuremark utilities available don't use any real gaming engine. If the benchmark doesn't provide a set of performance tests that are reflective of the graphics engines powering the games of today and tomorrow, then how would you expect said benchmarks to produce real results that you would see reproduced in those graphics engines being used in the games? Anyway thats all I'm gonna say on that, you could read a more detailed write-up on the problems with the latest benchmarking utilities Here and Here . You might also be interested in the controversy caused by "Optimizations" in both benchmarks and games that disable features you might prefer enabled which can be found here . And you can find information on one alternative performance testing method that is actuallly using a real game engine that people play games on here Here . Take a few minutes to read about all this and you might be surprised what all you find out. I'm inclined to think the FX5900's performance may not be quite as strong as what some benchmarking utilities might suggest.
This of course is entirely just my opinion as I don't own one myself (I'm quite happy with my ATI 9700pro and soon will be even happier with an AIW 9700pro too). Not trying to crap on Nvidia, but I do think that alot of the Questions being raised at Hardocp (did you know you can't use the shorthand [ H ] without having some HTML crap happen in these forums) are justified and valid and the answers to those questions deserve some scrutiny from the community.

 

[Lonyo]

Originally posted by: NYHoustonman
Huge drop in prices for the FX 5900 128 on pricewatch the past few days. It's apparently now down to $290 shipped retail box, from around $340 last time I checked. Anybody have any idea why, and if I should now consider this more seriously vs. the 9800 pro 128, which was what I was leaning towards because it was cheaper? Is it worth the extra $30 bucks? It was a done deal before this, but now I dunno. Is all this stuff about the 9800 being better with shadows going to affect performance in future games (mainly HL2) alot, or is it just a load of bs?

Well, maybe there's a link between the price of the 9800Pro going to $299 after rebate at CompUSA for a retail card?

 

[nRollo]

I think it's just of move toward MSRP. Aren't these cards supposed to be $300/$400, and retailers have just been jacking the price because people were willing to pay due to the absence of a new nVidia based card for so long?

As far as performance, I doubt you'd notice the difference between this and a 9800Pro. You wouldn't be able to go above 4X8X quality with this card, but you can only go above those settings in performance mode on a 9800Pro. (which gives you some iq issues) There are some who say the lower shader performance will affect future games, but they seem to have no answer for the superior D3 performance, and there is a total LACK of other game benches to see if they're right. Time will tell.

So I'd say it's pretty much a wash between the two, your preference.

 

[Pete]

Originally posted by: CVNet1
I wouldn't put much faith in benchmarking (ie 3dMarking) anymore as it has been shown that driver "Optimizations" can easily make a products score increase by a significant amount. And even discounting that, the latest Futuremark utilities available don't use any real gaming engine. If the benchmark doesn't provide a set of performance tests that are reflective of the graphics engines powering the games of today and tomorrow, then how would you expect said benchmarks to produce real results that you would see reproduced in those graphics engines being used in the games? Anyway thats all I'm gonna say on that, you could read a more detailed write-up on the problems with the latest benchmarking utilities Here and Here . You might also be interested in the controversy caused by "Optimizations" in both benchmarks and games that disable features you might prefer enabled which can be found here . And you can find information on one alternative performance testing method that is actuallly using a real game engine that people play games on here Here . Take a few minutes to read about all this and you might be surprised what all you find out. I'm inclined to think the FX5900's performance may not be quite as strong as what some benchmarking utilities might suggest.
This of course is entirely just my opinion as I don't own one myself (I'm quite happy with my ATI 9700pro and soon will be even happier with an AIW 9700pro too). Not trying to crap on Nvidia, but I do think that alot of the Questions being raised at Hardocp (did you know you can't use the shorthand [ H ] without having some HTML crap happen in these forums) are justified and valid and the answers to those questions deserve some scrutiny from the community.

Good Lord, stop referencing [ H ] as an authority on 3D! Stop the madness!

The price drops are interesting, but they shouldn't be about MSRP, as the 5900 is $400 and the 5900U is $500 MSRP. Also, nV will debut the 5700 at $300 MSRP, so there's really no place to go cheaper with the 5900 without impinging on the 5700. I ain't complainin' 'bout lower prices, though.

 

[JulesMaximus]

Originally posted by: CVNet1
I wouldn't put much faith in benchmarking (ie 3dMarking) anymore as it has been shown that driver "Optimizations" can easily make a products score increase by a significant amount. And even discounting that, the latest Futuremark utilities available don't use any real gaming engine. If the benchmark doesn't provide a set of performance tests that are reflective of the graphics engines powering the games of today and tomorrow, then how would you expect said benchmarks to produce real results that you would see reproduced in those graphics engines being used in the games? Anyway thats all I'm gonna say on that, you could read a more detailed write-up on the problems with the latest benchmarking utilities Here and Here . You might also be interested in the controversy caused by "Optimizations" in both benchmarks and games that disable features you might prefer enabled which can be found here . And you can find information on one alternative performance testing method that is actuallly using a real game engine that people play games on here Here . Take a few minutes to read about all this and you might be surprised what all you find out. I'm inclined to think the FX5900's performance may not be quite as strong as what some benchmarking utilities might suggest.
This of course is entirely just my opinion as I don't own one myself (I'm quite happy with my ATI 9700pro and soon will be even happier with an AIW 9700pro too). Not trying to crap on Nvidia, but I do think that alot of the Questions being raised at Hardocp (did you know you can't use the shorthand [ H ] without having some HTML crap happen in these forums) are justified and valid and the answers to those questions deserve some scrutiny from the community.

I've read a lot of the stuff you posted links to already. The benchmarks I am referring to are synthetic benchmards as well as the standard game benchmarks like UT2k3, Q3 and a few others. That's just my observation from reading through the reviews. Personally, I think it's crazy to spend $500 for the latest video card when you can get one for half that which will play any game out there admirably. I'm saying that the difference in performance between a 5900 Ultra and a 5900 non-Ultra is probably not going to be very noticeable in real world gaming.

I've always been kind of partial to nVidia card but with their high pricing and the poor performance of the 5600 series cards I decided to buy an ATI card (I'm not new to them-I have one in my wife's computer). It all came down to $ for me.

 

[godspeedx]

Originally posted by: JulesMaximus
From what I've seen in benchmark testing there isn't a heck of a lot of difference between the FX 5900 and FX 5900 Ultra in performance. Someone correct me if I'm wrong?

I'm not positive on this, but the reason you don't see much of a difference is probably because not many programs can really "use" all that the higher end cards have to offer. Once more advanced stuff comes out you will see a bigger difference.

that's what I think?

 

[nRollo]

but they shouldn't be about MSRP, as the 5900 is $400 and the 5900U is $500 MSRP.

You are right Pete. Must just be the market setting the price, with R9800s so reasonable, those MSRPs would result in very low sales for a comparable product. When I bought my 9800P a few weeks ago, I didn't see any benchmarks that made me think a 5900U was worth $80 more.