[SOLVED] Can core voltage alone make you crash?

culzone

Honorable
Apr 27, 2016
555
1
10,995
I have experienced some odd things with my core voltage lately.
im stable at 1.296v 5ghz on my cpu (did alot of tests for several hours)
However adding more voltage should give more stability right? if i wanted to be AVX stable for example.
If I go anything above 1.32v I crash within 10 minutes, and if I increase even more..
1.34v crash after 5min (ish)
1.36v 3min
1.38v 1min
etc.
Never going above 90C in any of the tests so should not be crash because of this.
Anyone experienced this before? The big question for me is why does it crash the higher core voltage i go?
The VRM heatsink is hot AF if i would've touched it for more than 2 seconds, I would hurt myself.
[Test used prime95 v26.6 small ffts]
Im using a z390 phantom gaming 6 mobo
9900k
2080 ti
DDR4 @ 4266mhz CL19
Corsair RM750x 750w PSU

All in a good custom loop 3x 240mms radiators
 
Solution
For a cpu to function, things need to be balanced. That means when you raise vcore, the cpu lowers current to match the power limit. Not enough current = unstable. When you raise vcore, it changes the vdroop, which with added LLC for vdroop lows, skyrockets actual voltage, lowering current = unstable. There's also other factors like vccio and vccsa which get messed with = unstable.

Out of balance voltages = unstable. Jacking up vcore doesn't make a pc more stable, it's either stable or not.

Karadjgne

Titan
Ambassador
For a cpu to function, things need to be balanced. That means when you raise vcore, the cpu lowers current to match the power limit. Not enough current = unstable. When you raise vcore, it changes the vdroop, which with added LLC for vdroop lows, skyrockets actual voltage, lowering current = unstable. There's also other factors like vccio and vccsa which get messed with = unstable.

Out of balance voltages = unstable. Jacking up vcore doesn't make a pc more stable, it's either stable or not.
 
Solution