This is regarding how VTTDDR helped me out with 3466MHz The Stilt
Link to ZIP
Organise files by time.
So changes from above txt was just frequency initially. ~32min in Y-Cruncher failed (4.075 1.287 3466S 0.943 1.37 Y FAIL room 23.1C.jpg). I proceeded to bump SOC by 1 step, less the 8min fail.
Kept bumped SOC and changed ProcODT to 60 and all CAD Bus 30, again fail, HWINFO in screenie shows 15min but fail was like <8min, I was away from PC.
I drop back to ProcODT 53 (ie what Auto was) keep bump on SOC and CAD Bus, test lasted ~24min. So defo need ProcODT 53.3, dropped SOC back by a step again as that bump hadn't helped, added 1 step VDIMM, fail <8min (4.075 1.287 3466S 0.943 1.375 Y fail.jpg).
So looking at what went on and what was used to pass HCI ~100% (~30min) I knew the settings I was playing with were not helping. As I knew CPU OC was fine due to testing and setting a step higher than tested there was something memory related I was missing on tuning up.
The answer was VTTDDR
, set to 0.687 = ~2hrs and counting Y-Crunching
. With the VTTDDR tweak all settings were as the HCI pass, SOC: 0.943V, VDIMM: 1.37V, etc, etc.
Nice to see you here
Damn, nooo! no TPM header :/ .