• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

New to oc trying on a 10900K

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Leoniums

Registered
Joined
Jan 10, 2019
Merry Christmas to all
I recently bought a 10900K with an MSI Ace mobo.
At the time i received my new hardware i saw that the cpu had thermal paste residues on it.
I called my supplier and waiting a reply. At the meantime i said.. why not lets try it and if its good i can keep it. I am on custom water loop and have 32Gb ram g.skill 4x8gb
Also new to overclocking.
I tried 5.1Ghz @1.29v which seems stable on prime95 no avx, testing small fft but temps going up to mid 80s. vccsa going crazy up to 1.38 (?) and vccio 1.3. Core vid 1.318
I tried 5.2 with 1.31 and 1.32 which is near to stable with core vid 1.338 but if i try to lower the vccsa i got errors (small and big ones).
Is this an a average chip? Should i keep it or replace it?
Opinions? Thanks
 
I have read tour article more than 10 times :)
And to be honest its a great article and i would like to thank you for that.
My ram is 4x8Gb 3600 cl16.
I have read tons of posts that vccsa shouldnt past 1.3 but not sure if my oc fails due to low voltage or due trying get lower vccsa.
Tried 1.29v @5.1 no avx offset all other settings to auto 70mins on prime95 small fft with avx no issues. Vccsa hit 1.385
I am trying again now @5.2Ghz no avx offset with 1.34v vccsa1.25 vccio1.2. Temps with ac on and a fan hitting my case.... temps from 80 up to 93.... running 20 mins now small fft prime95 (Windows freezed)
Are these temps normal?
I hit 90*C with 1.34v and you had 1.41v for 90*C
And what about this chip :eek:
 
Last edited:
iF YOU'RE TRYING TO ISOLATE THE ram THEN USE SOMETHING LIKE aida64, ha caps lock was on. Small FFT doesn't use the RAM at all so if it's failing you know it core voltage.
Work on one thing at a time and try not to change multiple items in BIOS. Dialing in the final numbers takes patience.
Your temps seem to be where they should be and the chip seems fine. They're not all the same if anything my CPU would be "leakier" that's why it takes more voltage, that plus my cooling isn't quite as good as yours would be. I have the 360 EK predator which is quite good but still an AIO.
 
Looks like core instability. I got that before, (total system crash) when there was a core wall.
 
Thanks for the replies and the info.
I am running a 480 XE radiator.
Updates
I tried again @5.2 Ghz with vcore (placed on bios) 1.335
LLC auto, vccsa vccio auto as well and uncore ratio 43
I run 110 min full prime 95 w/o any issues. Some cores hit for a while 94 but went back to 86 almost immediately. I got an error though (event viewer) for a searchapp or something... but probably is due to previously windows repairs.
Hwinfo64, vccsa was hitting 1.385 again, VR Voot max was 1.33v,
Right now i am trying again with vccsa and vccio 1.25 / 1.2, almost 1h no issues.
I think except the temps, the chip isnt bad and i should keep it.
If i pass the current prime95 run i will try 5.4Ghz with -2 avx offset i think.
Any opinion about the chip and if i should play with the LLC or leave it at auto? Should i try get higher uncore ratio ?
Code:
 
Last edited:
If you're going to drop the SA voltage you should test with something more memory sensitive like AIDA64 or even P95 on blend mode. BEFORE you try to go any higher with the core speed since you know it's stable
 
You are right i shouldnt had run again prime95 with vccsa vccio changes only. Although its running 100mins now w/o issues.
I will stop it and i will check memory stability or a prime95 w/o avx @5.4 -2avx while having vccsa vccio on auto again.
Will report back. Thanks again
 
I am trying prime95 blend 20 tortures, with unigine heaven at (ultra settings) as you suggested... Unigine video isnt totally smooth though... i dont know if this is normal
1 worker stopped.... no error though!
I am trying from 1.335v to 1.34v and checking again, or the worker stopped for another reason, like memory, gpu etc?
 
Last edited:
A other small update.
I tried 3h your combination, prime95 blend with hunigine heaven.
While i used ctrl+alt to jumb into the prime screen and see if everything is still running, windows freezed. Restarted jumbed into bios, freezed also there... restart and bios again.
Changed vccsa vccio value from 1.25 and 1.2 to auto.
Tried again your combination. After 8h no issues (overnight) i used ctrl@alt again to check prime95 workers and everything was running ok.
Opinions?
 
Sounds like you have a decent stable overclock to me.
 
A other small update.
I tried 3h your combination, prime95 blend with hunigine heaven.
While i used ctrl+alt to jumb into the prime screen and see if everything is still running, windows freezed. Restarted jumbed into bios, freezed also there... restart and bios again.
Changed vccsa vccio value from 1.25 and 1.2 to auto.
Tried again your combination. After 8h no issues (overnight) i used ctrl@alt again to check prime95 workers and everything was running ok.
Opinions?

I would call that stable. I don't think I've run a test program for 8 hours in years because I'm lazy and impatient. I'd call that very stable and run with it.
 
Changed vccsa vccio value from 1.25 and 1.2 to auto.
What are the actual values now? Typically when left on auto and using XMP, these are set WAY higher then they need to be. If it's above 1.3 and 1.2V, I'd work to tweak the down, personally.
 
Vccsa going up to 1.385. BFV still has stuttering... not like before but its there.
dont know whats going on to be honest.
Right now, went back to bios, load default settings, restart.
Pc doesnt start...
windows seems corrupted... pfffff
can the chip be bad? Didnt had memory issues before i changed cpu
 
Last edited:
Can the imc of this chip have a bad time with 4 x8gb and thats why it gives such a vccsa value?
I am pretty sure cant be an oc issue now... i just installed windows nothing more..
Gonna give a try with 2 memory sticks and see whats going on.
 
The best way to isolate a memory issue is to set the PC to stock enable XMP and then test it. AIDA64 is free for trial use and can find memory instability quite quickly. One more thing that can help with 4 sticks is bumping the DRAM voltage up to 1.37 V maybe a bit more a bit less whatever works'
 
Vccsa going up to 1.385. BFV still has stuttering... not like before but its there.
dont know whats going on to be honest.
Right now, went back to bios, load default settings, restart.
Pc doesnt start...
windows seems corrupted... pfffff
can the chip be bad? Didnt had memory issues before i changed cpu

So you went back to defaults? Don't try to fix Windows before doing that!
 
back to default... windows didnt repair.. never started.
Re install everything while on default bios settings... using 2 sticks of ram now.
Everything looks ok. BFV no issues here.
Tried to run with 4 sticks again. VCCSA still goes up to 1.36 even with default settings, even trying with a pre made memory of msi for compatible memory @3600 cl16
BFV starts stuterring again.

Install
 
Last edited:
Another update. Tried manual vccsa vccio 1.15 both.
BFV looks much better if not perfect. Most stuttters if not all have been eliminated.
So this vccsa vccio probably have messed up my system.
I wish i knew why the msi bios settings put it so high.
Gonna try more settings to see if i can get where it should be.
maybe i should loosen my timings a bit as well.
 
I wouldn't loosen the timings, just adjust the voltages until it's stable
 
Back