yeah after discussions with Roman, it seems more people had issues with the original TG block in terms of idle and load temps, and he mentioned it was due to the nickel plating properties, which from his video, they sourced another supplier of, and redesigned the fin density of the new block.
After polishing the die even more (as i was bored), and retesting applying more/less LM, my CPU now idles in the late 30s/early 40s which is to be expected and never goes beyond 72c on CB23, 55c on OCCT large dataset. All fans at 500rpm
That said, while -30 CO and thermal limit of 75 was stable during benchmarking, not so much during general use/gaming, with it often crashing, so i knocked it back to -28. Too lazy to do per core.. maybe when i have some free time.
I think i've repasted it enough to the point i'm not gonna get any better.. but think it could be related to the restriction of the loop.. with the number of QD i'm using temporarily, will see what temps are like when i go back to hard line.
