Radeon 9600 XT Review

Rys, good luck. With nV's avalanche of drivers of late you must have balls of steele to take such a gamble.

 

Well I prefer a more in depth review this time with a better comparison with a 9600XT instead of giving up... Of course, you can also change your original review as well..

About the driver issue: There are a lot of websites who constantly review each driver.. And it seems that the performance in these reviews is actually the performance the users will get. I do not know what they are waiting to release the driver now, but with each version I see same performance increase over 44.03 with bug fixes in the previous versions.. I guess they are still cleaning the remaining IQ bugs within some games so that they can make a flawless launch this time..

However, I do not how they will get a way with the fake tri-linear, (maybe as they will get away with the 3dmark scores as nobody is analyzing the quality as before in the new reviews )

 

I see the side of the story about not testing with unreleased drivers, but what option did Rys have? AFIAK, the only drivers that support the nv36 are the leaked 5x.xx drivers (or maybe the 45.33? dont remember). And as far as a comparison for which card you should get if you want to get said card in the first place, it is good to see them performing as they will when they ship. It is obvious now that the 52.xx series drivers are the best drivers for the fx series to date, and a new/public release is not going to change performance by much, if at all. All in all, I enjoyed seeing Rys' review (although it was early) due to it being the battle of the new midrange cards.

 

To simply not include the card in the review at all.

 

Oh that makes far too much sense, we know the drivers used are using the Pseudo Trilinear/Bilinear garbage while the 9600 is using full trilinear...++ 25% increase right there for Nvidia.
Reviewers are well aware of this, yet contine to put up the pretty graphs, why not even the playing field and benchmark ATI cards in peformance mode also...but that would be too 'fair'.

No IQ comparisons, no true DX9 tests...Halo 2, how about the most popular online game right now BF 1942.

 

Well actually, if you play the game, 70 fps is completely useless. It took me some time to get verification from my aquaintance who play RTCW at high level, but he (and his team-mates typically) uses com_maxfps 125, which means that he sets the maximum framerate to 125. In order to maintain that framerate solidly, the system has to achieve on the order of 300 fps on average.

On todays systems, compromises in image quality have to be made to ensure that level of performance.´

OK, back to bashing Rys for publishing interesting, if unverifiable, data.

 

I play my games for fun.

I don't want to turn off IQ for framerate. I don't normally play FPS games anyway, so for me 70FPS in a game like RTCW sounds perfectly reasonable for normal non-competitive play.

 

Are you suggesting that people who play competitively aren't having fun?
In games like RTCW all gameplay is competitive. If you're playing a weaker game due to your settings, you're a liability to your team-mates. Doesn't mean you have to care, though.
I simply wanted to contribute a real datapoint from someone who actually plays the game a lot.

 

Makes sense, given ATi's and nV's history, though ATi's superior performance in UT2K3 in OGL is an oddity.

 

Rys should have made the review a preview.
(With the preview text in large font size).


Joe DeFuria you have alot of valid points but personally i think your a bit hard on Rys. In my opinion its ok to do a preview of any hardware or game if you point out that it is a preview and not a review.
Paper magazines and tv shows previews everything from cars to mobile phones before it reaches the user market so why shoudnt that be done with GFX cards as long as it points out that its a preview of an upcomming product.

Lets just say i want previews as long as they arnt made out to be reviews.
This forum has alot of intresting people posting for example both Joe and Rys write intresting stuff to read, keep up the good work.

/Nevyn

 

LOL I can't speak as to whether those who play competitively are having fun or not, though I would hope they are.

I was simply trying to demonstrate that there is at least one segment of the gaming population that doesn't require the highest framerates, but rather wants higher IQ. For those of us in that group, the massive framerates desired by your competitive groups are less important. So for people like myself a card that can pull 70FPS in RCTW is perfectly reasonable.

 

I don’t mind the idea of a Hexus “preview.” Hardware sites always like to get a scoop. My criticisms of the Hexus preview stem from ATI’s benchmarking Whitepaper.

Really it comes down to this; the premise appears to me --- that Nvidia is … “optimizing” … (whatever that means) … virtually for only the 10 or 12 benchmarks/games that hardware sites generally use. And that these benches provide a very skewed picture of the overall performance of the 5600U/5700 compared to the 9600pro. Going by many of the various Hardware site reviews of the 5600U, one would think the 5600U is at least as good as the 9600pro (maybe even a little better). A good selection of benchmarks provided in the ATI Whitepaper, 25games (outside the 10-12 usual suspects), paint a very different picture. The 5600U falls considerably behind the 9600pro --- 40%.

For whatever reason; such as lack of additional games with easy to use benchmarks, by only using those 10-12 games with easy built in benchmarking facilities, hardware sites like Hexus are not providing readers with an accurate portrayal of the overall relative performance of these cards. In fact you seem to be way off.

EDIT: THat's would be Hexus not Hexas.

 

I disagree. RTCW has a nice single player mode that doesn't require such high framerates.

 

Excuse me Jo..

But was the 9500pro only 199$? Did the ti 4200 have faster ram and only cost 199$? I dont seem to remember things being that way... i remeber buying a ti 4200, the one i am using now for about 250$ and that was after it had been out for a bit.


Perhaps im wrong.

 

My 9500 Pro cost $181 online.

 

Errr I have them...... so ur mistaken they are availeble to the public.

 

has the 9600xt's clock speed been finalized. Seems like theInq (yeah, yeah I know) seems to be under the impression that ATI might increase speed to better compete with nv36.

later,
epic

 

Same here most important do is that minimum fps doesn't drop to low.........

 

500 core (set) 300 memory (not sure the memory is finalized)

 

Yes. (If memory serves, it was $219 with a $20 rebate from day 1.)

The 4200 didn't have the fastest ram of the day, but it (like the Radeon 9500 Pro) had the highest end GPU of the day, the GeForce4 Ti.

The 5700 may have fast ram, but it is not a high-end GPU, nor does it have the 256 bit memory interface that the high end GPUs have.