Question OC beginner question i7 9700k

[losinbruce]

Hi!

I have been trying to get a stable 4.9ghz OC on my intel i7 9700k and was wondering if I could get some help. I went for 4.9 since that is the clock speed I would get by enabling enhanced turbo boost on in my bios and I felt I could get away with a lower core voltage.
So I started off by setting my core voltage at 1.310 with LLC 4 and override ((manual) C-states and speedshift is enabled) on my MSI Z390-A Pro which according to HWINFO went up to 1.320 maximum and 1.312 at full load (Lower LLC, LLC5, would drop the vcore down below 1.300 during stresstest and LLC3 would put it way higher). This voltage made realbench (2.43) stresstest stop prematurely as instability was detected. I upped the voltage in my bios to 1.320 which went up to 1.328 maximum in HWINFO and would be stable at 1.320 at full load during realbench stresstest. I ran the stresstest for 41 minutes where the message "Result Hash Match! and "Result mismatch. System Unstable. HALT!" would appear once. But the stress test didn't halt on it's own. I came back to check on it after 12 minutes or so and for half a second both of my monitors went black and came back on, maybe that's a sign of instability?

My temps during the stress test were:

Core 0 - Max 85 C
Core 1 - Max 84 C
Core 2 - Max 89 C
Core 3 - Max 87 C
Core 4 - Max 88 C
Core 5 - Max 86 C
Core 6 - Max 84 C
Core 7 - Max 84 C
CPU Package - Max 89 C
The highest average was on Core 2 I think and was something like 82 C.
I have a Noctua NH-U9S and I don't know the exact temperature of my room, but I guess it's around 20C which makes me feel that these temps are a bit high especially because I set my noctua to 80% speed when it reaches 80 C.,.

I have seen people suggesting letting the stresstest run for 8h up to 24h but with these temps and the messages I got in realbench made me not want to leave my computer unattended for so long..
Are these temps normal and could you help shed some light on what the realbench messages mean? I guess "Result mismatch. System Unstable. HALT!" is self explanatory, but I am a bit confused as to why realbench sometimes halts the stresstest on it's own and in another instance gives me that message.
I did run CoD Warzone both at core voltage 1.310 (which went up to 1.312) and 1.320 which both seemed fine. Though it was in practice mode and not online with 100+ players so I guess the load on the CPU was way lower than in a normal game.

I have not fully grasped vdroop yet. But since my vcore drops 0.008v from idle to full load, am I right to assume that is vdroop?

I understand that a stresstest puts way more load on the CPU than I would normally achieve. I use my computer mostly for gaming with weekend sessions stretching up to 8-12+ hours and some video editing in Davinci Resolve.
So as long as my temperature stays at normal level (which usually are around 60-70C with short peaks in the lower 70Cs) and I don't get any BSODs, would it be perfectly fine to leave everything as is? Other signs of instability would be a normal screen freeze right? And if any of those things occur I would just have to increase the core voltage or drop the multiplier right?

Thanks in advance!
Erik

 

[losinbruce]

UPDATE:

I dropped my OC to 4.8ghz because I wanted more manageable temperatures.
After testing different core voltage starting at 1.250, I ended up with this:

VCORE bios 1.280V Max VCORE recorded by HWINFO 1.288V VCORE and under load during the stress test it stays at 1.280V. During the test I again got these messages in realbench: 6 "Result has match!" messages and two "Result mismatch. System unstable. HALT!"
Stress test passed
Timer 30 min

Core temperatures during the stress test:

0 Max 79 AVG 72
1 Max 78 AVG 71
2 Max 82 AVG 75
3 Max 80 AVG 73
4 Max 80 AVG 75
5 Max 80 AVG 73
6 Max 77 AVG 71
7 Max 77 AVG 72

CPU Package power Max 122.506 W AVG 110.500 W

Now the two messages of "Result mismatch. System unstable. HALT!" worries me because obviously it's not completely stable according to realbench. But the stress test ran the full 30 min without ending prematurely with the message "instability detected" like a lot of the tests did before I ended up setting the vcore to 1.280. I know the test isn't optimal since it only ran for 30 min.
So my question again is should I assume that this OC is stable for my daily use and that I should only be concerned about setting the vcore higher if I encounter a BSOD or a computer freeze?
I would also like to have my vcore on adaptive so it drops during idle and I've seen people suggesting to leave CPU core voltage mode on "override" and instead changing windows energy alternative to "balanced". But this doesn't seem to drop my idle vcore..
If I change voltage mode to adaptive will it overshoot my set vcore by alot on an i7 9700K and would that mean I need to set an offset to counterbalance this so I can stay on my desired vcore? Will this offset mess with my vcore when the cpu is idle?

Thanks in advance!
Erik