Quote Originally Posted by MusicIsMyLife View Post
To be honest I am shocked to see "reviews" done with dry ice cooling.

It's well known that Phenom II scales with temperature. Doing such an marketing move I would do everything to make sure everybody who receive one piece will get the most out of it - of course with ln2. That TWKR is a special thing for the top dogs and it should be presented like it is - and not with clockspeeds that an Intel-user can achieve with watercooling.

Who is impressed to see 4.7 ghz with dry ice? I'm not. 6+ ghz and we are talking (no matter what cooling is used). So I don't get the point why AMD send such cpus to people who don't have full access to ln2.

That's no offense but I have seen lots of things from AMD that aren't well done (especially in Germany). And due to the fact that people got pieces and didn't test with ln2 I'm a bit disappointed.

I almost forgot: Great clocks and scores around there. Congratulations to all who are involved in this event.
A few things to add.

First, the thermal conductivity of pure copper actually increases from just above the LN2 to LHe and below range. This is a remarkable coincidence creating the opportunity for lower temperature and thermal conductivity with LN2 and LHe. So, as you state,a high static dissipation processor that can scale will shine most brightly under the ultimate test of LN2 and LHe.

However, I suspected that a well-designed DICE pot could, in the right hands, bring out some of the goodness of a TWKR. As long as you can shed the heat, you have the benefit of higher current potentially driving higher clocks. And as many of you know, chew* has produced outstanding DICE results in the past, and came through with a 5.35GHz 4 core validation and a 5.13 Pi run. That's a hell of a result and shows that a properly tuned DICE rig and a more efficient thermal interface have potential with high static dissipation processors.

We will see more results; these parts have been out in the wild only for days and arrangements for a proper LN2 run require some time. This is the beginning, not the end. We are simply spoiled by chew* and a few others who were ready to fly with these parts like a bat-out-of-hell. I know of a few events coming together that will continue this experiment...

Sincere thanks for everyone's comments. This is a challenging time for our industry (and just about everyone), and we appreciate the encouragement. We don't mind the criticism. Just remember that AMD is not a colossus, it is a group of people mostly in Austin and Markham, with a manufacturing partner mostly in Dresden, working to push the limits for the few while serving millions.