Originally Posted by ESRCJ
Here are my timings for 4000MHz. You have some headroom with your secondary timings. The highest tested vdimm by manufacturers is 1.5V, as they sell some kits at that voltage. So that should be safe for the most part.
Nice timings, quite tight, a way more aggressive than mine (all of them), you think they can safely work on a daily setup, not a benchmark one?
Let's say I'll start trying them, which of them should I start with?
I currently have 1.45v training and 1.40v all the rest so I could do 1.50v training / 1.45v windows.
Originally Posted by sblantipodi
Is there anyone who trust intel that pushed 18 cores to 5.1GHz with a standard cooler?
It sais "standard liquid cooling" that's standard custom water cooling like we all have (mostly), not LN2, water chiller nor a mere AIO.
Originally Posted by Jpmboy
Cache buys a lot in the AID64 memory test. 3.3 is up there... what is the actual Vcache with a 230mV offset? Remember, the offset is added
to the VID (stack) value.
1.45V VDIMM on the x299 platform is a safe zone on the 79xx series chips. I had my Omega at 1.5V VDIMM (4200c17 with 64GB) for several months... until the 9900X CPU died. Not sayin it was the VDIMM... just that it was at 1.5. The ram sticks are fine. I also have a pair of Royals running at 1.6V 24/7 on z370/8086K right now... no "adverse" effects yet on that platform.
Agree to 1.45v, can I still use a higher training voltage?
About the cache, I was keeping it for like 6 months at +0.230 offset at 3.2Ghz now it's +0.200 which is about 1.12v under hwinfo64, Asus EC, Cpu Cache/Mesh (1.119 - 1.122).
AIDA cache and memory test added (@ +0.200), over 4h stable
Ran at +0.180 and there was instability in half an hour.