While I'm wary regarding the potential IPO returns of Bitmain, their current portfolio shows them running out of their 16nm BM1387 chip-sets. These 16nm S9's are becoming long in the tooth, but as they use TSMC as their silicon foundry, and TSMC has a fabulous 7nm process in production, I expect it's only a matter of time (months) before we see a 7-5nm version of the S9, that would likely set new records in the J/TH area. The die shrink's main benefit is you can typically clock the devices faster with lower net power consumption. The costs of developing this new process is considerable, so I would expect part of this IPO would be to enable this transition. Several operations (like GMO) have gotten here already, but I believe once Bitmain shows up to the 5nm party and given the breadth of their mining portfolio, they would could well be a $3B+ operation next year. The only down side of this IPO is there are going to be some early Bitmain founders cashing out, which is always tough to accept.