**Official DFI LanParty UT X48-T2R(and LT) Discussion

[Rinaun]

Praz, how would you suggest to move the gtls together (0/2 1/3)? one ata time, or in sets? I just can't seem to stop one core from failing. as soon as I get close to correcting the gtl offset, another core fails.
So far I've deduced:

Cores 0 and 2 are 2nd setting, 1 and 3 cores are first.
NB GTL has nothing to do with the other gtls.
in Prime95, core 0 isthread 1, core 1=thread 2, core 2=thread 3, and core 3= thread 4
please Correct any if wrong.

I've also noticed some patterns, such as if you have the setting too far one way, if you look closely at P95 and which cores fail, you can figure out which way to go with which gtl.

I set up an experiment. I started with 137/124/112 gtls at 1.360 VTT. with this setting, core 1 would fail. For every time I went PAST the stable point using the second setting(in this case 127), the OTHER core would fail. For example, if thread 1 was failing, when I raised that cores GTL control (thread 1 = core 0) past the stable point, 127, core 3 would fail instead. I tried again with 125, and sure enough, core 1 failed this time. Then I tried 128, and got a fail on 3 again. I'm wondering if this is just pure luck, or the actual way GTLS work and an easier way of adjusting them.

edit: Also, is there any link to a guide on taking those measurements from the motherboard? I'm a bit timid to stick my fat hands down there. Any ways to make it easier to read?