User Manual

ASUS Z390 Series Feature Manual

It’s worth noting that AI Overclocking does not adjust the AVX instruction offset due

to limitations associated with this parameter. AVX workloads generate more heat and

typically require lower clock speeds, so if you run software that uses those instructions, we

recommend manually adjusting the offset value to compensate. The same algorithm that sets

the all-core frequency also predicts the maximum stable AVX frequency, so it’s easy to check

the prediction and modify the offset appropriately. This is also covered in the step-by-step

guide included in the UEFI.

Man vs machine

To see if I could do better than our algorithm, I disabled the auto-tuner and tried to reach

higher speeds than the 5.2GHz it recommended. This time, I booted back to the desktop for

software tweaking with the latest version of AI Suite. I rst tried to dial up the multiplier by one

step, taking it to 53X for 5.3GHz, but the system locked up. Incrementally increasing the base

clock to 101MHz got me to 5.252GHz without issues, but that’s only about 1% higher than

what AI OC achieved, and it took a lot longer. I tried to push things further by adjusting the

voltage but didn’t have any success.

AI Overclocking is an amazingly quick tuning utility that goes beyond standard automated

overclocking tools and remains vigilant in the face of changing system needs. It’s a handy

weapon for beginners looking for an easy path to higher speeds and a useful tool for experts

who want to get to a stable baseline with minimal fussing, but it’s only available on select

boards in the ASUS Z390 family. Check out our introduction to Z390 gaming motherboards

(https://rog.asus.com/articles/maximus-motherboards/introducing-rog-maximus-strix-z390-

gaming-motherboards/) in the ROG Maximus and Strix families, and don’t miss the full

ASUS lineup in the complete Z390 motherboard guide (https://edgeup.asus.com/2018/

z390-motherboard-guide-coffee-lake-8-core/) at Edge Up. We have boards for everyone from

beginners and rst-time builders to hardcore enthusiasts and prosumers, and most of them

support AI Overclocking.