Hmm, seems like I got 2.3-->2.5 to work.... not entirely sure since I have no clue what this rig did on normal smp before.... no bigadv to compare times too, and it seems down on power from the wall but that could be bigadv-->smp as well.
if thats the case, maybe I'll try for an undervolt.
So, cant be sure of the freq, but the undervolt had some results. went from 1.1625 to 1.0 and the (regular smp) full load went from 610w to 575w, could be a difference in wu. But idle draw went down 10w. just gonna give it a few wu before I call it a success.
Ok, Charile, you definetly want Sfield/Tear's modified TPC....Original only handles 1socket. I loaded it with just the -l flag and it showed that only the first 12cores were adjusted by the non modified version. So I used the modified one and wow! Idle down another 20w and load down to 480w (1.0v vcore).
holy crap but whats the frequency they're running at?
I set them to 2.5, but if that only affects ES chips then they're running at the stock 2.3, its known that the TPC changes arent "visible" with other cpu info tools in linux, they report what the chips settings tell them it should be. From what I've read the only way to be sure is run some benchmarking at stock and at changed values and compare. F@H is my benchmark.... albeit useless without bigadv :(
I was never much or a linux guy but I remember DOS being a whole lot easier command line wise.... I still dont understand what I'm doing, just following others and googling commands others glance over cuz they're basic.