MotherboardsReviews

ASRock X399M Taichi – AMD Threadripper Motherboard Review

Overclocking

The 1920X processor is rated at 3.5GHz with max turbo up to 4GHz. The X399M Taichi is able to make it run in boost mode up to 4.17GHz on single cores for short period of time. It’s more than we could expect and at the same time fully stable.

Default voltage ID of our processor is 1.225V. Depends on load, at auto settings it goes up to 1.45V what is maximum voltage specified by AMD and at the same time maximum safe voltage specified for longer work.

We wanted to keep as high frequency as it’s possible on all cores using the NH-U14S TR4-SP3 cooler from Noctua. Below is what we could achieve.

4.15GHz was possible for benchmarks but couldn’t run without issues under longer full load. I guess it’s just matter of cooling. To achieve stability, the frequency had to be lowered to 4.1GHz and voltage to 1.4V. It’s a great result and even better than most AM4 Ryzens can do.

While trying to lower the voltage, I’ve decided to try maximum frequency at 1.3V. I was also surprised with the results as I was able to set 4GHz on all cores without any problems. I guess it’s optimal for daily work and runs cool and quiet using even lower Noctua coolers.

Since I found the maximum frequency for the CPU then I’ve moved to memory overclocking.

Memory affects almost everything on AMD Ryzen and the Threadripper is not any different. As I already mentioned earlier, I was able to set Patriot Viper 4 memory at DDR4-3733 using nothing but XMP profile. Trying to set more, I was able to run tests at DDR4-3866 which is nearly maximum possible for the Ryzen. However, at DDR4-3866 I couldn’t set anything fully stable. I guess it’s because of the platform as anything above 3600 is really hard to stabilize, regardless if in use is a processor for AM4 or SP3r2 socket.

Maximum stable frequency using Samsung single rank modules was DDR4-3733 so XMP profile on mentioned Patriot memory or manual settings at a bit tighter timings. Maximum stable frequency using Micron D-die was DDR4-3600. I had no memory based on Hynix IC that I could test in quad channel mode. There were no issues to set DDR4-3733 in dual channel mode on Hynix C-die.

A bigger challenge is always dual rank memory. For tests has been used Ballistix Elite DDR4-3000 memory kit which is based on Samsung B-die. All 64GB of RAM could work stable at DDR4-3466 CL16-18-18 and 1.35V. Pretty good result considering it’s quad channel mode.

ASRock made everything really easy. Regardless of used memory IC, all works great. There were no issues with memory profiles on six different memory kits. Also, all tested drives and other devices were working as expected.

 

Related posts

Leave a Comment

* By using this form you agree with the storage and handling of your data by this website.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More