HardOCP's review of X1800XT - ambiguous?

I’m trying to make heads or tails out of HardOCP’s review of the X1800 XT - and to be honest… I’m finding it kind of ambiguous. On the one hand they show NVidia’s latest hardware leading in most of the benchmarks, but then they state theis in the summary

In some games, the BFGTech GeForce 7800 GTX OC pulled out ahead in terms of performance. In others, the Radeon X1800 XT came out on top.

But looking at each of the benchmarks closely(at least the way HardOCP tests) NVidia’s hardware was the clear winner. http://www.hardocp.com/article.html?art=ODc1LDE=

Have any of you tested or worked with either card yet? If so, what are your feelings about (it)?

How do you say that? Probably “ex eighteen hundred ex tee,” but I’m going to go ahead and call it the “exty-eighteen hundred ex tee.”

HardOCP doesn’t like to do apples to apples reviews, which makes it quite difficult to determine which hardware is actually better.

Instead they compare the radeon at 1600x1200 with 2x adaptive antialiasing and 8x anistropic filtering to the nv at 1600x1200 with 2x transparency multisampling antialiasing and 4x anistropic filtering because they feel that each is the “highest playable setting”. Modern hardware has so many different features that it can be very difficult to puzzle your way through… wtf is transparency multisampling antialiasing, exactly? Is it worth the performance hit? I doubt even case knows without looking it up, and he does this shit for a living. And to make it even more confusing they put the two different configurations on the same graph!

I disagree with their reasons for this. I can read the charts and figure out which is better but many others can’t. Different, as they say, isn’t always better. It’s nice to know what hardocp feels the highest playable resolution is on their specific hardware setup but it doesn’t really affect my purchase all that much. I want to get the best value for my money, and for that I need a nice clean apples-to-apples comparison.

Anyway…

The numbers are summarized correctly in the conclusion; the cards are basically neck and neck, but for now the radeon is louder and more expensive.

Thanks Stusser - you hit the nail on the head. Glad to know it’s not just me.

The noise issue does concern me.

The X1800 XT is definitely faster than a NON-OVERCLOCKED 7800 GTX. Especially with AA and AF enabled.

It’s also more expensive, and pretty much out of stock everywhere. And don’t expect Nvidia to sit around and let ATI ship a big expensive 512MB card without response.

Then there’s this little tidbit:
http://www.extremetech.com/article2/0,1697,1880749,00.asp

I find ATI’s new programmable memory controller and tread dispatch unit interesting. I wonder how much faster the cards can go with driver updates?

Transparency AA rocks. It’s what makes trees get to be antialiased, too. And fences, or whatever else has avoided being improved by AA since the move to MSAA a few years ago. I found Supersampling transparency AA to be spiffy in WoW.

HardOCP reviews, though, have become very close to worthless – and their proximity to that point increases with each new reader who purchases an LCD for gaming. I couldn’t be much less interested in someone else’s idea of a “perfect” resolution/settings balance if I’m playing on a fixed-resolution-or-it-looks-like-ass LCD. Their “max playable” information could be derived from any decent set of benchmarks at a glance, but good luck getting useful information out of their graphs unless they happen to magically choose the resolution that matters most for you. Too bad, too — they used to be worth reading first. Now they’re often not even among the reviews I read about a new product launch.

Jason, I’d been watching that too but to be honest. I just don’t see vid cards having the extra overhead to do anthing more than they already do (in games that is). Outside of a game there’s some good potential.

Do you think we’d be able to add 3rd party cooling solutions to it?