[Warning Fudzilla] GeCube HD 2900XT beats 8800GTS 640MB

Tom|Nbk · 4 Jul 2007 at 14:11

An updated review from fudzilla, fair enough he is known for his speculation etc, etc and criticised for it. But this is a fair review done with both 7.5s and 7.6s shows the card in a much more positive light against the GTS.

http://www.fudzilla.com/index.php?option=com_content&task=view&id=1744&Itemid=40

Tom|Nbk · 4 Jul 2007 at 14:20

helmutcheese said:
Ahh well the GTS 640 is not the top Nvidia card.

Yea but the 2900 was never going to be a GTX competitor

Tom|Nbk · 4 Jul 2007 at 14:25

bfar said:
It should have been

Suppose but ask yourself which sell more the GTS or GTX,

Tom|Nbk · 4 Jul 2007 at 14:29

Dutch Guy said:
Maybe not now but I am sure that a while ago that was the idea, until they got problems with the cores.

Looks liek AMD did a lot of driver optimising, especially in Company of Heroes, but F.E.A.R. still shows a bigger drop in framerate when enabling 4xAA where the 2900XT's score is more than halved

Yea perhaps if they had done another respin it could have been a winner, but I guess they had already lost enough cash designing it and couldn't afford it or justify it, maybe they think they should have now.

Tom|Nbk · 4 Jul 2007 at 16:02

Dutch Guy said:
They had a choice of delaying even longer or go with what they had at the time and they chose the latter.

I think the large performance drop is because AA is done in shaders and if it is with a game that relies heavily on shaders the drop is bigger than a game that isn't 'shader heavy'

Precisely.

Tom|Nbk · 4 Jul 2007 at 18:30

esiemi said:
well done ATI, nearly a year late and still there is no solid proof of 2900 beating GTS, and even if really beats GTS what about the heat/fan/noise/AA etc issues.

Multiple reviews show it to beat the GTS :confused:

heat/fan/noise is not really THAT much of an issue not to me and most of the other owners on this board anyway.

Tom|Nbk · 4 Jul 2007 at 18:54

Uriel said:
They don't show some of the games the 2900XT was really struggling with, like STALKER.

That benchmark was confirmed a load of **** that showed STALKER being bogged down, runs just as my GTS did here

Tom|Nbk · 4 Jul 2007 at 19:34

Cyber-Mav said:
you need to show both sides of the story, no point posting loads of threads showing 2900 beating gts in certain situations, it just makes it sound like your trying to justify your purchase of the 2900

Post made no sense but no change there, I can fire up S.T.A.L.K.E.R right this minute and performance is never as low in those reviews, Im not trying to justify anything. Just the truth.

Tom|Nbk · 4 Jul 2007 at 19:45

Cyber-Mav said:
so then why post links to reviews that show good performance for some games yet bad in stalker, and then you blame the benchmark they do? its like selective reading.

you abviously have something on your shoulder since you have started so many threads to do with ati and the 2900 that its quite ridiculous. really not sure what your trying to achieve here

You shouldnt trust reviews Mav. Not trying to achieve anything infact your the only one that has a problem with me, seems like your the one with the chip on your shoulder.

Tom|Nbk · 4 Jul 2007 at 20:00

A lot of reviews use different setups/configs/drivers you just shouldnt trust them as they can be very varying, for once I would have thought you would have at least tried to engage your brain to understand that but no all hope is lost

. CBA replying to you anymore Mav, your impossible.

Tom|Nbk · 4 Jul 2007 at 20:51

helmutcheese said:
This is my view in my short time here and I am not going to add to this or bait arguements:

There is 1 person in general who is a nice dude but has made many threads about the new ATI card he has, I read into it he is trying to convince himself he did good buying the card as he was on Nvidia now ATI then was going to do a swap back to Nvidia with another user and now thinking about buying Nvidia until again decides to stay with the ATI.

I personally would put these cards down to same as FX5000's and they could actully game ok but overall were dudes, skip the ATI range this time as I did with Nvidia, I went from GF 4 TI4600 to 6800Ultra, has no hard time gaming inbetween.

I was kind of same with my AM2 over C2D but I made my own decision on variables and will leanr to live with that.

Ok, I'll stop making threads about the card I own, since when have you seen a person that owns a nvidia card make an ATi thread anyway, also the fact I am a beta tester means I get a lot of the information first hand. NO body on here is totally unbiased. But Ok even though it creates good discussion I'll stop and give myself a slap on the wrist.

Tom|Nbk · 4 Jul 2007 at 21:03

helmutcheese said:
I just think the more you torture yourself the worst it gets (IMO), being there with "did I get right kit etc" many times.

I've made one, yes one thread about getting a GTX as I wanted to step up past the GTS/2900 level, nothing about regretting purchasing it.

helmutcheese said:
You can do whatever you like, you seem a nice guy, beta testing means little, I did same for MS and I'm no expert on software.

When you speak to AMD daily about what the 2900 has in store and get defined answers about its ARC it means a lot, well to me anyway

helmutcheese said:
If the card runs game the way you are happy thats all that matters, I have others on my case as I have Raptors and a 8800 Ultra, you dont see me telling anyone not to buy them if they can afford.

It does. Depending on how some of the bigger releases perform on this card later this year will I know that im keeping it for the long run. If they perform how I want then great.

Tom|Nbk · 4 Jul 2007 at 21:09

Cyber-Mav said:
no don;t stop altogether, thats not gonna solve anything.
all i want to know is why you post links to reviews then tell me not to believe them.

i didn;t say anything to you about stop making new ati threads. so don;t even go there.

i just want to know why you say a review which shows benchmarks of the 2900 beating the gts is a legit review but the second the review shows the 2900 being beat you then say not to trust the review.
iv read all the reviews you have posted and it just seems to me that you say reviews are good where the 2900 is the winner and in that same review turn the page it shows stalker or some other benchmarks where the 2900 is the looser you make it sound as iff only thing to trust is benchmarks that show the 2900 beating the gts. any benchmark that shows the gts winning is not trustworthy.
hmm you can see why i get a suss as to what you say.

how can i put it, there is being biased, there is also being fanboy, now the new extreme is being a TomNbk

The fact that S.T.A.L.K.E.R is a very unpredictable game also. God knows why they even bothered comparing with that (one of the shoddiest coded games in a long time) maybe even worse than R6V Mav

. FPS was up one minute and down the other with both my GTS/2900 it all depends on the scenes being rendered in game and a nuber of other reasons. This is all im trying to get across don't trust reviews with S.T.A.L.K.E.R as the benchmark, I think you didn't get the full point I was trying to get across. :rolleyes:

Tom|Nbk · 4 Jul 2007 at 21:19

killer_uk said:
It was, but that doesn't mean anything. All it means was that the coding is less efficient. Stalker would be very hard to bench and get reliable results across the board due to it's design. The game is a big world in which different events can happen, thus making performance very unreliable if you were to benchmark it.

Thank you you put it into much better words than I could be bothered

Cyber-Mav said:
depends on exactly how much cheese is on your helmut

:/ tbh.

Tom|Nbk · 5 Jul 2007 at 22:58

Im not actually sure why they chose to resolve the AA in the shader units. I believe there is an explanation somewhere I'll try to find it.

Tom|Nbk · 5 Jul 2007 at 23:04

willhub said:
He must be using a crap CPU or something coz in 03 I get 52,000, 05, I get 17500 and 06 I get 11442 which beats the 2900XT on there.

But on the game benchmarks, it beats the GTX once, and some its not far ahead of a GTS, there on Par I would say, both nice cards, anyone will be happy which ever they choose

Did you have your GTS clocked when you did those benches?

Tom|Nbk · 5 Jul 2007 at 23:07

willhub said:
Yea at 650/2.2Ghz, at stock I get like 10200 in 06, I think thats with my CPU and ram at stock too.

Why on earth would you compare your clocked benches to a review which was done with stock cards :confused:

obviously you'll score higher.

This is the testbed

Motherboard:
Nforce 680i EVGA board (Supplied by EVGA)

CPU:
Intel Core 2 Duo 6800 Extreme edition (Supplied by Intel)

CPU Cooler:
Akasa EVO AK 922 Blue Athlon 64/X2/FX cooler and Intel CPU's (Akasa)

Memory:
OCZ Reaper PC2-8500 1066MHz 5-5-5-15 (Supplied by OCZ)
CL5-5-5-15-CR2T at 2.3V

Graphics card:

GeCube Radeon HD 2900XT 512MB GDDR3 (Supplied by GeCube)

Leadtek WinFast PX8800GTS 640MB (Supplied by Leadtek)

PSU:
OCZ GameXStream 700W (Supplied by OCZ)

Hard disk:
Seagate Barracuda 7200.9 500GB SATA (Supplied by Seagate)

Tom|Nbk · 5 Jul 2007 at 23:16

willhub said:
Those 3DMark 06 scores are too low for that setup at stock if I get 10200 at stock...

Nah the GTS's score around 9k @ stock, take Loadsamoney for example he has everything stock and just scores 9k.

Tom|Nbk · 5 Jul 2007 at 23:20

Will go for it. If you put your system at stock you will have the same score as him as he has the same CPU + GPU.

Tom|Nbk · 5 Jul 2007 at 23:24

Here we go from when I had my GTS.

http://service.futuremark.com/compare?3dm06=1705723

GPU @ Stock CPU @ 3.4Ghz

9745

If you put your cpu to stock youd be looking at just 9k. See?. The Fudzilla results are bang on and until I see a compare URL from you stating otherwise......................