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

10900k OC on Asrock Z490 Taichi

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

KWEEK

New Member
Joined
Jun 19, 2021
Can someone help me get a stable OC. I've been trying to hit 5.1Ghz with issues. Just looking for a starting point and some numbers I can enter into BIO's to start stress testing the CPU for stable OC. I can't find any good info online about OC with this Mobo. If someone can shoot me into the right direction that would be great. I just reset all my settings in BIO's so I'm ready to start again with some help from the forums.

Thanks!

Specs:

10900k
32GB 2 STICKS G.SKILL 3600 16-19-19-39 2T
Asrock Z490
RTX 2080 super
1TB 970 EVO
 
Last edited:
Hey Kweek!

So... two sticks of RAM or four? Start with the RAM not using it's XMP (so 2133) regardless.

As far as a starting point, Try 1.30V static/override LLC1 or 2 (whatever keeps your voltage under load as close to what you set in the BIOS - IIRC, when you set a static voltage it automatically sets to 1), and stress test. If it passes (maybe 15 mins), lower the voltage 0.05 and try again. Rinse and repeat until it's not stable, raise it back to whatever was stable before and test longer. Once you get the CPU dialed in to reasonable stability (a couple hours stress testing), enable XMP for the RAM and see if it's still stable.

What issues have you run into, exactly?
 
I was just getting BSOD during a cinebench 20 loop test.

I have 32GB and two sticks

Should I run something harder on the CPU like prime95?
 
With 32GB, you may need to play with some voltages to get that stable. So, set the RAM back to auto so it sits at 2133, then work on the CPU.
 
Could you guide me on what settings to change. Sorry I’m new to asrock mobo and don’t really understand what settings to adjust.

I’m in bios right now under OC tweaker

I’m seeing CPU Config and voltage config

Should I set cpu to all core and 51 then set voltage to 1.30 fixed voltage, Then set load line calibration to level 2?

I notice on voltage config at the top I have options for voltage mode: stable and OC mode. Just keep it at stable mode for now?
 
Last edited:
Thank you, I will try some runs right now and get back to you.

I noticed I’m seeing a maximum voltage of 1.447v is that safe or normal? Under load it’s at 1.365 and with no stress test it’s always around 1.4v

This is on all the cores under core 0 Vid using HWinfo64 where I’m seeing these numbers
 

Attachments

  • E811F5FA-B6A7-4BCD-BAB8-4AF6FE0E5BEC.jpeg
    E811F5FA-B6A7-4BCD-BAB8-4AF6FE0E5BEC.jpeg
    347.6 KB · Views: 5
CPUz is showing 1.288v

I noticed when running small FFts prime95 I'm hitting 94c then BSOD. This is with 51 and 1.30v and Level 1
 
Last edited:
That's better. Use that to see load voltage (idle doesn't matter much).

90C is fine anyway so since you're gaming, you'd be all set.
 
Maybe! You need to test it. Every CPU is different. ;)

What I gave you is a starting point. It may (should) be stable for gaming... it may not be. Test and see. The goal is to reach your overclocking speed using the least amount of voltage. Maybe you can reach 5.2 GHz... who knows... but you need to do the work to get there. :)
 
what's the best stress test to use for prime95 blend or small fft's?

I wanna do a test that's more realistic workload. Because I know prime95 pushes the limits for the most stable but I'm just using this for gaming and never going to push the limits like others do when Overclocking. any other programs you recommend for a more realistic workload?

also I notice my physical memory load is hitting 99% when running blend on prime95, is that normal?
 
Last edited:
Small fft is cpu only. Blend is memory and cpu (note p95 says that...you select the test amd it has a description).

Normal for a memory test to almost fill the memory.

You can loop cinebench... there's several others out there and everyone swears by them soooo......
 
I just notice during a cinebench20 loop I'm seeing a windows hardware error (WHEA) total errors: 1

what could be causing that error? it pops up after a few runs of cinebench
 
Last edited:
I just ran a small fft's test on stock settings with xmp on. I don't get any BSOD or crashes but these temps seem so high. Might be time to reapply some new thermal paste? it's been almost 2 years since I've replaced it and I used the stock paste that came on the NZXT Z73 AIO cooler.
 

Attachments

  • Capture.PNG
    Capture.PNG
    102.5 KB · Views: 2
I think I got a pretty good stable OC on 5.1 with 1.3V

I did get 1 error at the end of the loop on cinebench 20, I was reading around and afterburner could be causing this error? Could that error be a issue or don't stress over it?

I did notice errors when running prime95 "CPU Cache L0 errors"
 

Attachments

  • NEW.PNG
    NEW.PNG
    21.9 KB · Views: 3
  • Capture.PNG
    Capture.PNG
    439.8 KB · Views: 2
  • 2.PNG
    2.PNG
    6.9 KB · Views: 3
  • error.PNG
    error.PNG
    41 KB · Views: 2
I've never seen that hardware error before....we don't use that app for that. Does cidnebench fail after a couple hours? P95?

Keep the cpu under 100C for stress testing. ;)
 
I’ve done 20 passes on cinebench with ram running at 3600mhz and at 5.1Ghz on the CPU with no BSOD or crashes. Should this be stable enough for gaming? my highest temp was 96c


I'm scared to run prime95 because I know the temps will go over 100c for sure after awhile. I'm happy with the results just from cinebench because I know the games I play will never hit that kind of stress on the CPU ever

I blame the high temps on my house temp. it's 80 degrees right now in my house
 

Attachments

  • 1.PNG
    1.PNG
    336.6 KB · Views: 3
Back