The hardware industry has a love/hate relationship with (competitive/extreme) overclockers, we all know that. On the one hand, overclockers are the group of people most passionate about the depths of the product they sculpt. BIOS, hardware, compatibility, VRM, you name it - everything is closely looked at. On the other hand, it's also the group who is most vocal about potential problems of hardware. As one Product Manager working for a motherboard vendor puts it: "Overclockers are the [annoying] 1%, but they push us to improve our products every generation." Perhaps one of the most tiniest of details was recently discussed on the forum: achieving 102.9 MHz BCLK on locked Skylake processors, giving one vendor in particular an edge in the XTU 2xCPU rankings.
That vendor is ASRock. Based on the information provided in xxbassplayerxx's forum thread, it appears they dominate the XTU 2xCPU global rankings. For those who are not in the loop, there are two important things to know about this ranking. First of all, it is dominated by the locked Core i3 6320 Skylake processor which has a multiplier of 39x and is limited to approximately 102.9 MHz base clock frequency when overclocked. Secondly, the XTU benchmark is heavily affected by the lack of AVX instruction support when unlocking the overclocking capabilities of the Core i3 processors (more information here). That means the ranking is effectively capped by the limited base clock frequency. Now, here is where it gets interesting. When looking at the top-100 best results in the XTU 2xCPU category, 85 (!) results use ASRock. Quite remarkable considering that ASRock sits at about 12% market share of Z170 according to HWBOT statistics. Even more incredible - no other brand has representation among the results that score higher than 740 marks (ASUS, GIGABYTE and EVGA have minimum one entry at 740).
In a plea to the motherboard vendors, xxbassplayer writes "We need to petition the other manufacturers to unlock this! It ain't pretty for them." Whether the plea will see any response is doubtful. After all, this is (yet another) problem of a specific group within the 1% and affects no real-world applications or situations.