I've seen comparisons where xmrig ran without large pages or where tester was recording a video on the same pc while testing xmrig's hashrate, now this is what I call unfair My equipment is pretty standard, I just make sure everything is tested in perfect conditions - properly tuned, no background programs eating CPU, updates and unneeded Windows services turned off etc. I need this to easily reproduce the same hashrate at every test - it's hard to tell if some change in the code improves hashrate +0.1% or makes it worse -0.1% without this kind of stability.