ASUS strix z390-e, 8700k, basic safe overclocking help needed for a first timer. Concerned with VCCIO/VCCSA.

Oct 28, 2018
12
0
10
Sorry for the long post but it’s a long story and I'm very new to this, want to include as many details as possible.
I’ll start by saying I recently built my first computer and would like a simple safe overclock. After inquiring on multiple forums I’ve gotten conflicting responses to my questions which led me to making slight changes that ended in scary results for me.

My system components:
-ASUS ROG Strix Z390-e
-I7 8700k
-32gb tridentZ RGB 3200mhz ram (more than needed but it was a gift)
-Coolermaster ma620p dual tower dual fan air cooler (massive)
-850watt 80+gold Thermaltake PSU
-ASUS gtx 1060 6gb (I don’t pc game but may try getting into it)
-phanteks Enthoo pro case
-2x 250gb Samsung 970 evo m.2 ssd in raid 0 for boot drive
-1x 500gb 850evo ssd
-1x 6tb toshiba hdd

So what I originally did was :
-enable XMP_II profile and said no at the prompt
-Set core ratio to sync all cores
-set the multiplier to 47 as i just wanted a safe basic overclock
-I left everything else on auto because I have no idea what I’m doing and want to keep it safe and long lasting.

Max Voltages per HWmonitor:
Vcore: 1.314v
VCCIO: 1.344v
System Agent: 1.216v

Everything has been fine with these settings and I’ve been running them about 4-6 months.

But Someone replied to my previous reddit post about the same topic and said my vccio and vccsa were borderline dangerous and to lower them to 1.1 to 1.125v.
Personally I thought that was a drastic amount to lower them by.
So what I did was lower them slightly.
If I remember correctly
- I changed the vccio from auto, which was reading (1.344), to manual and set it to somewhere between 1.2 and 1.3 as opposed to the reddit suggested 1.1v.
- I did the same to vccsa, set to manual and just lowered it slightly to 1.2 from the auto setting which was reading like 1.22-1.25.
- I F10 saved and tried to reboot.
—THEN—
-***My system (which is basically brand new) began an endless loop of trying to start, couldn’t post, and I wasnt even able to get into bios to change anything back as it just kept restarting; No amount of F2 F5 or delete spamming would let me enter the UEFI to change settings back.

-What I ended up doing (because I wasn’t too sure about how to use the cmos jumper or what it even does) was remove the small watch style battery, wait a little while, plug everything back in. And restart.
- After 2-3 cycles of the computer attempting to start it finally did and I was able to get into bios where everything seemed to be set back to stock.

-At this point it was very late and after a scare like that I just left it all alone except enabling the xmp2 profile so my ram would run at 3200mhz.
- Other than that I’m back to square one with all stock settings and honestly a little scared to make any changes.
-Originally I was very happy with my first all core 4.7ghz OC with everything set on auto. Until someone said it was dangerous to run vccio and vccsa at such high voltages and that I needed to dial them down right away. Which I attempted to do and failed miserably but thankfully didn’t brick anything.

So my ?’s are :
1. What did I do wrong? Did I lower the vccio to much to fast?

2. Why Did my system go into endless cycle of trying to boot and not be able to get into uefi?

3. Did I have to remove that battery or could the cmos jumper have fixed my issue?

4.**side note the ”memok” *switch is set to on by default** Should I switch it off when doing changes like this or leave it on?

5. And lastly, since lowering those SA and IO voltages was clearly an issue, would it be ok to set everything back the way I had it, with the only changes being sync all cores to 47 and enabling xmp2?
Are those IO and SA voltages really too high? Or can I just set it back the way I had it and be happy with a safe and basic OC.

Thanks in advance to anyone willing to take the time to help a noob such as myself.
It will be greatly appreciated.
Sincerely, Paul Jolly. 
 

kasol kay

Honorable
Mar 22, 2013
232
0
10,760
This Video goes over all the important settings one by one on an Asus board (it's a Z370 Board it might slightly different to yours):

https://www.youtube.com/watch?v=pbRauc8gpQM

As far as I'm aware 1.3-1.35Vcore on Intel Chips is relatively safe as long as your Temps are under control.

You can set the multiplier as high you want obviously how stable it is depends on your Voltage and CPU, so you could even try 48-49 on the Multiplier.

Maybe someone with the exact same Chip can answer your question with more expertise but I honestly think that a Voltage of 1.3-1.35 sounds very safe as long as you have adequate cooling.
 
Oct 28, 2018
12
0
10
Thanks for the reply. I wasnt really concerned with the vcore voltage as that seemed totally fine and I’m not having any temp issues. More concerned with the vccio voltage of 1.344. And even after setting it to manual of 1.3125 and then even lower to 1.3 it still reads 1.344 in hwmonitor.
I also downloaded hwinfo which has a ton of readings. Some if which have to be wrong or just reading nothing... Like Tsensor 1or2 reading 8C. And tsensor 9 reading 106C after a cold boot with my computer located in a non heated room in dead of winter. It’s soo cold in that room that I wear a jacket when going into it. What could those readings possibly be. I even set everything back to completely stock settings and hwinfo Tsensor9 still reads 106c no matter what. Super weird and less important. More concerned with my auto and manual vccio voltage being 1.344 no matter what i do. Or should I not be concerned as everything has been running fine for a while and I was only concerned after a redditor said “that’s dangerous lower it immediately”. Since then I’ve had other commenters saying that it’s just fine. Soo many mixed reviews.

Thanks for the help and insight guys. It’s greatly appreciated for a noob like me just trying to learn.