I believe FC gave the market the proper opportunity to absorb the product before this change in direction. I also believe he made it quite clear that this was a possible outcome. My preference is to see a sustainable business model that reinvests in maintaining a sizeable percentage of the network as equipment sales margins will only diminish unless there is a profound rise in the price of BTC or something profoundly cracks the current ASIC cost/efficiency situation to be highly profitable for a short term. That term being only as long as it takes for the competition to step up or the manufacturer elects to use their own equipment.
I suspect shares may be making their way to steady hands with what appears to be a longer term strategy then some had believed. Holding ground with a percentage of the network in a position of profitability appears to be a favorable option when you are benefitting from working at the most competitive costs. AM has deep enough pockets to achieve this and is already in possession of ASICs to make it happen.
i think you are sorely mistaken here.
Maintaining a tangible network percentage as well as selling individual miners is an enormous undertaking when you consider that every 6-8months your miners become obsolete. The amount of associated cost with designing and manufacturing new boards every 8 months is enormous. So is shipping internationally and having to deal with thousands of customers who need tech support.
The best model is what they tried-- oem sales of chips. It shovels off all the responsibility to procuring pcb boards, design of such, etc. to a 2nd party. This is preferable because they can concentrate on what they do best-- best access to foundries and ability to push chips to the market fastest and quickest. The overhead with this is by far smallest, and the points of failure of delivering a product to your customers is smallest; either your chips work or they don't. No need to design and wait for boards, powersupplies, cables, shipping materials, etc. As long as they have a strong design team down to 28nm, this is far far far preferable.
Yeah but if this fails and the demand isn't strong enough or maybe your chip even underperforms (gen 3), you need to resort to the things you can still do. And self-mining or franchised mining seems to be the best option now. There just hasn't been enough demand for gen 3.