Question Is this overclock possible on the i9-9940X?

[Servs]

https://www.userbenchmark.com/UserRun/15514920 < Current PC Specs

I want to push 2 cores to 4.2Ghz. The CPU cooler is a Corsair H150i Pro AIO. I know its gonna get hotter, but my question is can i even hit that high of a OC on an i9-9940X?
New to OC btw.

 

[celrod]

Hi!
I'm also new to overclocking, and happen to have an i9 9940-X, and the same watercooler (Corsair H150i Pro AIO; 360 mm radiator).

My motherboard is an ASUS TUF X299.
RAM are 4x (to take advantage of quad channel memory) Ripjaws V Series DDR4 PC4-25600, 3200 MHz 16GB.

I updated my bios, turned on X.M.P. (there seems to be only a single profile, or disabled) and set the overclock to "fast".
This set the base clock to 4600, and the AVX2 and AVX512 clocks to 4200.


I don't game, but spend most of my free time writing (hopefully fast) statistics software. The OS is Clear Linux.

Someone posted on an online forum BLAS: Julia slower than R; when trying their test (calling a couple of OpenBLAS's LAPACK routines from the R vs Julia programming languages) , I realized that my overclock on the i9 is unstable.
Before updating my bios, the computer would freeze and shutdown more or less the instant I started the benchmark.
The old bios was from May 2018, while the new one dates February from this year.

After updating the bios, I can run the benchmark in Julia without a problem. (It takes <8 seconds, while my Ryzen Threadripper 1950x takes close to 18 seconds; LAPACK is very AVX[-512] heavy).
CPU temperatures rise to about 60C within a second, stay there, and then drop to about 30C as soon as the benchmark is done.

But, the computer still crashes instantaneously when starting the benchmark from R.

What's the best way to increase stability?
If temperatures aren't that extreme, it sounds like I should increase voltage? Or load line calibration? I didn't see any options named "load line calibration" in the ASUS bios.
Or would it be better for me to change those default clock settings?

I haven't noticed system stability issues from workloads such as compiling software (all core, non-avx) or from running avx512 heavy single core benchmarks. The all core avx512 seems to crash instantly, or not at all.
Although, I didn't run things for very long (ie, 8 or so seconds), but temperatures remained stable over that time.