meh, i give up. that last core just can't get to 4.1ghz
1.4125 cpu vcore with medium llc, puts it up to 1.42v (minimum for 4.0ghz stable) i've tried 1.464v (both with medium LLC and 1.36v+extreme LLC for reaching that voltage) but aside from hitting my thermal limit i still can't get 4.1ghz stable under IBT
1.4v nb core (tested up to 1.45v still fails 4.1ghz)
1.65v ram when i OC it to cl10 2133, (tested 4.1 and 4.2ghz at 1.5v 1866mhz cl9 xmp1 settings and even 1600mhz cl11)
1.20v HTlink (tried 1.3v, still unstable)
1.8v NB/PCIE/PLL voltage (tried increasing to 1.9v, still unstable)
2.5v cpu pll voltage (tried upping it to 2.65v still unstable)
1.3v NBvoltage (tested 1.45v, still unstable + more heat)
those are all the voltages i can control via BIOS.
also those are the voltages required for 4.0ghz @ FSB 229, 2520NB, 2980 HT & 2133mhz cl10 ram
stress tested with all the power saving features (and even svm), along with turbo turned OFF while running the stock xmp1 cl9 1866 ram speed (hell i even tested for 4.1ghz with cl11 1600mhz ram), all my fans running at max speed, p95(v27.7.1, can't find 27.9) fails after a few minutes at 4.2ghz, IBT fails second or third test at 4.1ghz.
@4.1ghz IBT fails the second or third test, yet p95 passes (albeit only tested 20minutes), at 4.2ghz p95 fails too. tested both multiplier and FSB OCed for 4.1/4.2ghz, same volts as above, same results.
now i've enabled the power saving features (disabled turbo, HPC and APM) for 24/7 4.0ghz, 229 fsb and ram OCed to 2133 cl10.
pity, i should've bought another cheapo mobo, wish i had known this chip was so worthless before splurging so much on this motherboard. i've never let the chip hit 60C (and only bought it in november), the socket hit 61C while testing 4.1 but tctl is still under that threshold (max 58C)
think i'll end up swaping the 8320e for my phenom II 925 and OC it again. but first ill test to see if i can get 4.4ghz again off the first 6 cores like the 970a had achieved.
edit: 6 core only goes up to 4.4ghz stable, same issue as above, going to 4.5ghz fails IBT but works with p95
side note: possible my last core isn't being cooled properly by the separated 212+ pipes, although i doubt that's the problem
another note to self: leave the voltages @ normal, use dynamic dvid to make use of power saving features.
Edited by ocyt - 4/16/17 at 5:59am