I have had thousands of gears run under my hands, mainly from these three brands (Bitmain, MicroBT, and Canaan). If I were to be fair, I would exclude Canaan because I bought most of those used, and they were a complete mess. With that said, MicroBT was an order of magnitude more reliable than all the Antminers I had, with the exception of some variants of the S9 (not counting the S9k).
It's worth mentioning that the results of my experiments, which lasted for years at very different scales, were also done at very different places with completely different environments. We had small to medium-sized farms that were a complete mess with cooling and other farms where miners ran in perfectly air-conditioned facilities. The results didn't change much,MicroBT still beat Bitmain at every single level.
However, to be even fairer, if you were to focus on a single component, which is the PSU, then MicroBT vs. Antminer would likely be a draw. However, when it comes to the control board and hashboards (the most expensive and hardest to replace component), MicroBT has always been better. I literally had some old M20s that ran with a single fan while the heatsinks were completely covered in dirt. Nothing but an Antminer S9 would go that far.
It's also worth mentioning that with Antminers, it's a game of luck. Different batches from different manufacturing locations give you different results, so you could get a good batch today and a bad one next month. After all, it's Bitmain. Does anyone recall the 6-month failure report from the large Chinese farms back in the day, when Jihan Wu had to apologize and send his technicians on-site to fix those miners? Was it 30% or 40%?
|
