The Anandtech article focused more on 32-bit gaming performance (bad) while the TechTV article focused on the 32-bit applications they could (or couldn’t) get to run.
It would be unfair to say that any measurements made at this point in time are indicative of what the final product will be.
However, it would not be unfair to say that this product is in no shape to be heading out the door anytime soon, and that people shouldn’t expect this until September at the earliest.
Chicken and Egg
On the one hand, you might think that if AMD priced its Hammer more reasonably, a lot more units would be sold, and there would be increased pressure not only on Microsoft, but on hardware and software makers alike to get the load out and get cracking on AMD64 programs and drivers.
Let’s face it, you can’t exactly say the heat is on when ATI can’t even be bothered to come up with drivers at this point.
On the other hand, it could well be that AMD has been told by these folks not to expect Windows and driver support for AMD64 soon, and this is a big reason why they aren’t trying to ramp up Hammer now.
Whichever it is, unless you want to play beta tester, this unreadiness reinforces the notion that it might be a good idea for the average person to wait on Hammer until Windows is ready for prime time.
If September or later ends up being the release date, that more or less coincides with the release date of something else: 90nm Hammers. Even if pricing is no more reasonable on those than they are with the current crop, you certainly ought to be able to go further for your considerable outlay with a 90nm processor at the beginning of its ramp than with a 130nm processor at the end of its ramp.
Being an unpaid beta tester wouldn’t be so bad if you could at least run some things faster than you can with regular Windows, but paying a bunch now just to get to run them often rather slower seems more than a bit masochistic for anyone but those who truly get joy from this.