New 4.2GHz overclock that I stabilized yesterday, 21x200MHz at just 1.360V in CPUZ (1.376V actual). Here is a screen of a run I did with it this morning. Decided to open the balcony and let some fresh air in. Ambient temps around 17C.
64C on the hottest core using LinX (Linpack) 20-pass run with my new cooling setup running a customized CoolIt FZE with a Swiftech MCR220-QP rad and micro res in the loop.
One strange thing I noticed is that I have about 10C-12C discrepancy between the hottest and the coldest core when I load all cores at 100%. I tried reseating the block several times, changed TIM around a lot, shook everything multiple times to remove air over the last 3 days, topped off the pump as much as it can take, everything I can think of...still, can't bring the hottest core closer to the coldest core. Maybe 5C would be reasonable, but 10C-12C seems a bit excessive for a temp difference.
Any ideas why this would be the case? I feel like if I can solve the core difference problem I could drop another 4C-5C off the load temps.
Reply With Quote
Bookmarks