Question Z390 UD overclocking 9700K to 5GHz, constant BSODs/freezes without clear reason

[indynt]

Hey all, was wondering if anyone had experience with overclocking on this specific board, because I feel like I'm missing something. What's happening is that whenever I run prime95 it crashes within 3 or less minutes, while there is no apparent reason for it to crash. I've checked and played with BIOS settings, VRM temps, CPU temps, CPU voltage, etc.

CPU voltage I've currently set at 1.360V and I have tried incrementing up to 1.400V, which honestly just felt excessive for a 5 GHz overclock on a 9700K. When it crashes, usually the VRM temps are pretty low, sitting comfortably at 70 C, with the CPU usually sitting at 80-85 C. Ring clock multiplier is set at 45x (I've tried setting it higher, but it just kept freezing the windows startup sequence), and core clock multiplier is set at 50x, with a base clock of 100 MHz. Enhanced multi core performance and settings like VT-d (along with all the voltage managing settings) are disabled.

XMP is on, my system has 2x16GB 2133 MHz corsair vengeance sticks running a 3600MHz profile (I honestly don't know a lot about memory overclocking or XMP so I'm not sure if a 1500 MHz increase on memory is stable, but I've never had problems with it before).

Do you guys have any tips on how I can nail down the exact cause or information on the board? I'd love to hear any advice on this matter.

Note: The problem does not seem to be the memory, I used y-cruncher and prime95 to test it and encountered no freezes or crashes.

 

[Ralston18]

Update your post to include full system hardware specs and OS information.

Include PSU: make, model, wattage, age, condition (original to build, new, used, refurbished)?

Disk drive(s): make, model, capacity, how full?

Any error codes, warnings, or informational events in Reliability History/Monitor or Event Viewer?

Will the build boot into Safe Mode?

 

[Eximo]

If this is an old overclock that has become unstable, CPUs do degrade over time, especially at voltages above 1.35 volts.

Single core was 4.9Ghz. All core turbo stock is 4.6Ghz, so 400Mhz across all 8 cores is going to add up to a bit.

Low to mid 80s is about right for daily use, but lower is better of course.

When you say all the voltage management settings, does that include load line calibration? If you are using a fixed voltage you might be experiencing voltage droop that is causing your instability.

LLC can push the voltages above where they are set, so best to be careful with it. But I would say lower the voltage back down to 1.36, bump up LLC if it isn't running now, and see if that helps at all.

AVX on or off testing with Prime 95. AVX loads do quite the number on the CPU, generally not needed if you are testing for gaming stability. Did you do any AVX offsets? Might be worth it.

 

[indynt]

If this is an old overclock that has become unstable, CPUs do degrade over time, especially at voltages above 1.35 volts.

Single core was 4.9Ghz. All core turbo stock is 4.6Ghz, so 400Mhz across all 8 cores is going to add up to a bit.

Low to mid 80s is about right for daily use, but lower is better of course.

When you say all the voltage management settings, does that include load line calibration? If you are using a fixed voltage you might be experiencing voltage droop that is causing your instability.

LLC can push the voltages above where they are set, so best to be careful with it. But I would say lower the voltage back down to 1.36, bump up LLC if it isn't running now, and see if that helps at all.

AVX on or off testing with Prime 95. AVX loads do quite the number on the CPU, generally not needed if you are testing for gaming stability. Did you do any AVX offsets? Might be worth it.

It's not an 'old' overclock, after years of stock usage, I decided I wanted to get some extra performance. Loadline calibration has been set to Turbo. I'll try to mess around with LLC. My AVX offset is set to 2.