hidde663

4.675~4.725 all core

"Your chip died because 1.55v is way too fucking much for daily" I tried to kill it, that was the idea, to see the degradation process. anyways hopefully someone will read this who is new to OC and wont try such a high vcore as a daily driver.

Too much? I don't know. The chip was delidded, resealed, 85C@155 watt. did not gain anything past ~1.536v on air, the extra is just for the droop. So i set 1.55v, did a few x265 4k benches in a row, which was the hardest test to pass in my book, stabalized the highest freq. without destabalising the mem. It ran fine for a few weeks, then suddenly after i came back from work and it wouldnt POST no more. (i did shut it down myself, so no crashes) Thats why i think it died due to a voltage spike at startup. (similar to how Jayztwocents 7700k died, cant remember for certain) I was hoping the degradation experiment would have lasted a little longer, but oh well got a 5800x now, which is quite a jump in performance. I will wait for de8auer's degradation vid before trying anything "un-safe" on this one ;)

random info: i killed this chip last month, with only 1.55v (AVX stable@4.924), no degradation - just instant failure. Boot voltage spike maby?

Nice!!

I'm sorry to say but, version 3.3.3 is not allowed: https://hwbot.org/news/11973_application_145_rules/ You can probably just re-submit as gpupi 3.3.3 1B

~5% gains from memory timings, i should be able to push to 1690/2260

congratz!

25mv less gives better OC headroom in timespy for some reason

Benchmark tessellation load modified by AMD Catalyst driver, result invalid. Check your video driver settings.

im not gonna lie, it took me multiple days just to tune the GDDR5, without it randomly spitting errors after an hour

trying again for 16300

one of the sticks now spit memory errors @2133CL17 :(

I know the same tricks bios mod, trixx for trixx

nice!

I can no longer get 2666cl14 to post, 2400CL16 does :(, the other stick still does 3466CL14, currently running my 2x4@3018CL12-15-15-28-1 + the 8GB B-die stick.

i've tried everything, tried to boot about 100 times, that one stick just wont post past 2666 anymore. high VCCIO/VDIMM doesnt help. but the stick had issues before, where it would not post 3200CL12, unless I trained with 1 stick and posted once, reinsert 2nd stick and then get into windows. and yes it makes no sense haha. I'll attempt a bios flash later today, im not getitng my hopes up tho.

1 stick degraded form 3200CL12@1.5 to 2666CL14...

Changed Tfefi to 65535 from 32767 after this and my 2nd stick will no longer POST, nice

nice matched leeghoofd on the ms haha

its starting to get cold here, i might be able to do 5600 graphics score on air ;)

baseline (24/7 stable)

baseline(24/7 stable)

for some reason i was unable to sign the cert to bmrbt64.sys, getting error #577 (unable to sign certificate, fatal error). been trying to fix/bypass this it myself for about 2 hours(probably some service i disabled or corrupted OS files) anyways bypassed it by disabling secure boot from the bios... (clean install of windows would also have fixed it most likely) anyways, happy to join the benchm8 era

Insane score for water!!, 1.85 for VDIMM@3800CL11?