Your score is high.
Your score is high.
I observed he was running a gaming pro that matches an M3, which might be typical. I’ve seen some YouTube tech channels compare the M5 and M7 from MSI, noting differences in voltage because of power phases on the board. The M3 is lower than both, so it could be close. With voltage regulation now handled by the boards more, the impact becomes more noticeable, which is why I selected my M7.
The easy tuner is included in the UEFI on the Asus x99 deluxe, functioning as an automatic overclock feature inside the BIOS. It tends to be quite cautious, capping performance at around 4GHz. I understand what you're referring to with OC software—back in the day I accidentally damaged a Q6600 and a 680i board using it. Now I avoid anything that requires afterburner or similar tools.
Sorry for the delay. Here are the images: my OC and realbench scores below. The setup: Asus Maximus VIII Hero Alpha with i7-6700k (stock 4.0, turbo 4.2 - manual tuning at 4.8/1.296V). I also have an Asus Strix R9 390 8Gb OC with DCIII (stock GPU:1050/vram:6000 - manual tuning at 1100/vram:7000, which was the highest frequency I could achieve with Asus GPUTweakII; it won’t let me go higher). For reference, check the bottom of this post.
Regarding your system, as you boost CPU voltage, heat output rises. These newer CPUs are generally more stable than older models. If temperatures climb too high, the system may throttle to protect itself—this can cause temporary slowdowns. In extreme cases, it might shut down if overheating occurs. After a few minutes of restarting, you should be safe again.
I follow this process for tuning: run Cinebench, Realbench, and 3DMark (full package from free version takes about 30 minutes). Then perform an Aida64 stress test (CPU, FPU, cache, memory, GPU only if you’ve already tuned). If all pass, there’s a high likelihood Aida won’t fail—usually it fails within the first 20-30 minutes. For extra confidence, run Aida for 24 hours.
Always perform BIOS-based tuning rather than using third-party tools that may be unreliable. They often give inconsistent results.