To be technical, ASIC miner CHIP manufacturing on the 14/16nm node is in either Taiwan (TSMC fab with input or part-ownership by Samsung) or New York State of the US (Global Founderies, ex-IBM fab) at this point.
Ah yes, I remember this from when I was researching mobile-CPU friendly PoW hash function designs in 2016, that Samsung had an advantage on 14nm chips for mobile.
Bitmain in particular has shown a LOT of issues with getting enough chips to make miners with- when you have LARGE companies like NVidia and AMD having availability shortages on their new cards that they plan to sell millions of, a small company like Bitmain has to settle for what capasity is left over (and keep in mind that AMD has long-term contracts locking in capasity at Global Founderies due to the ex-AMD fabs being part OF Global Founderies dating back to it's foundation and their sale TO GF by AMD - and a lot of the rest of GF capasity is locked into IBM for the SAME reason).
So doesn't this favor my hypothesis that Litecoin's price rise means ASIC supply is constrained thus hashrate will not rise as fast as price does, and thus older 28nm stock should come into mining and it should signal SegWit, because those guys want to remain profitable at higher prices.
I see that there are still a couple of KNC Titan Scrypt miners for sale on ebay. And as the price is rising the A2 Dominators are becoming viable to mine with. I see a shit load of A2s for sale on ebay at really cheap prices ($200, etc). You can practically but them for nothing and when the price rises enough you could flip then for a profit I bet. I haven't actually run the calculation to see at what price the A2s are generating enough profit.