[Ripper3]

Brilliant reviewing as usual, Wiz.
Those Q4 results are peculiar, but I noticed them in another review you did, the GTX280 ones I believe.
Speaking of the 280, I really can't see Nvidia selling many of these, especially if the die shrink in a few months is true, and considering the performance-per-$ compared to the 4850. Glad to see ATi fighting back.
I've been thinking of buying a gaming laptop for Uni, if I go, but at this rate, I'll just build a semi-portable PC, with a 4850 or 4870 inside.

Noticed the Performance-per-$ chart though, somewhat confused as to how the 3850 managed to get to the top, before I checked prices. Some American retailers seem to have them around the $130 mark, we have them at, the cheapest, between £75-80. I can see how it topped the chart there.

This is all good news for ATi. A card that is already kicking ass and taking names, without needing another driver set to get to full speed, but here's hoping 8.7s eek out something extra either way.

Edit:
Jelle Mees, I do kind of agree with you, but the point of reviews has always been to compare performance in demanding games. If it can play the demanding games, it's basically inevitable that it will perform everywhere.
Still, I do wish they used 1680x1050 in testing, but 1600x1200 should give very good indication of performance at the widescreen version.
What I really want to see is testing LOW resolutions with HIGH AA/AF, and overall high settings. Some people take advantage, lower the res to get higher settings, and AA allows them to counter the jaggies that appear more with lower resolutions. Although it doesn't happen as much with LCDs, since the LCD will generally display images a little rougher at non-native resolutions, and people want to avoid that happenning.

Boy, I need to pick a side and stick to it.