Iron

On Sunday and Tuesday (yesterday) we, the AOC, played with our phenny 1090T and the crosshair 4, to take confidence with amd since those are our first runs.

Let's start, setting up everything with 1090T, C4F, 8800GTX and Kingston 2133C9 Kingston provided us for a review: testing to find maximum vga frequencies on air:





Phenny @ 4,25 and 8800gtx at his max on air:


8K cpu score on air, not bad!
Then we started to do insulation on Ryba/sf3d gpu pot, to find max frequencies for the 8800gtx without still vmodding it:













Gpu pot mounted, here we go:






We arrived at -120 on the 8800, but I have an old UEI DT200 that's going crazy, so take these temperatures with a grain of salt...
At -120, we managed to close at 783/1620/972...


Contact between pot and gpu, good!

Iron

Given that we had problems, troubles, removing that kneaded eraser stuff from the Giga UD7, I tried a new way of preparing the mobo for the neoprene... the thought is to make capacitors and phases, "squared":




Done!


Neoprene...






And here we go with gpu and cpu under ln2:






After 2 hours, insulation did it!


So 5,6ghz max on the cpu, over 10k of cpu score.

http://hwbot.org/community/submissio...marks?new=true

Iron

Third and last part (for now )



11.159@cpu score, 6ghz stable with "only" 1,8v real, with cpu load calibration ON in bios.
We can't move on from that frequency by entering in win from the bios, we think we should try to enter @5,95 and then going up by using amd overdrive...

Last score with even the gpu@ln2, now we absolutely need the Vmod

A question: when we were@5,94ghz, the mobo was switching continuosly the frequencies by 60mhz (2mhz of fsb with 30x multi gives 60mhz)... that's not good that, when you are at approx 6ghz, the frequency goes up and down by so much!
Have you experienced that?

By the way: bios 0801, we had also a little problem with that NB voltage bug that Leghoofd was speaking about in his review for madshrimps

eva2000

sweet thanks for sharing

awesome insulation effort

unfortunately no experience with AM3 or LN2 so not sure about frequency fluctuations you experienced.

what NB voltage bug ?

Iron

Hi George, thanks!

The NB voltage bug is something that, when you have for example 1,3v, resets the voltage to 1,08 or 1,1v... So you lose time to understand "why this crap no more can enter into Windows while a few minutes ago it was 3DM06 stable" and then you notice with the readout points on the mobo pcb that voltage problem.
We worked it around by raising abnormally the nb voltage, so instead of settin 1,3 to have 1,25 real or so, we were setting 1,4v in the bios to have 1,2v real
Leeghoofd already reported it to asus, so probably in the 0901 bios they'll get rid of it