Quote from: minerpart on January 27, 2014, 07:02:35 PM

Two points:

KnC are not 'at 20nm'. KnC are working on a 20nm but that ASIC technology is still unproven. There is no guarantee that the reduced die size with limiting non-scalable components will allow for any significant increase in overall efficiencies over top-end 28nm. (Like the ACtM eASIC full custom 28nm). Infact, the increased cost of producing the KnC 20nm chip, plus the inevitable delays in actually designing and achieving a workable 20nm die (extremely difficult) could make KnC's Neptunes unable to ROI. That is a real possibility.

Second point - foundry capacity. Let's see some figures about the ability of the foundries producing bitcoin ASIC's being able to keep up with 100% production increase per month. No matter what their capacity now they are unable to meet that future demand - fact.

UMC currently has a huge capacity for 55nm but their total 28nm capacity seems to be only around 25% above that. So with a huge demand (100% increase per month) for 28nm, 28nm production limits will soon be met.
http://www.umc.com/English/class_300/c.asp

What that says is that very high demand for 28nm could reach a foundry back-log in the near future. That's why chip runs are indeed limited by the foundries - they do not have endless capacity when the world of electronics of all types runs off their production lines. 28nm is in high demand. 55nm is an established tech but with huge infrastructure supporting it and it does not have a 100% increase in market demand per month.

'Yau Kae Sheu, senior director of UMC's 12-inch Specialty Technology Development division, said, "UMC first taped-out customer 55nm SDDI products in late 2012. Achieving 15 million accumulated shipments and mature production yields in just one year is a significant milestone and highlights our engineering and manufacturing strength as a worldwide foundry specialty technology leader. With our 300mm Fab 12A in Tainan, Taiwan and Singapore's 300mm Fab 12i both providing ample capacity support and established economies of scale, we look forward to bringing additional 55nm products to volume production as many new products are scheduled for the design-in stage in early 2014'
http://www.prnewswire.com/news-releases/umc-surpasses-15-million-shipments-for-customer-55nm-sddi-chips-239014911.html

So the addition to our portfolio of a custom, top-end, 55nm that has a very high volume production capacity from UMC could be a very very fortunate - or smart - development for VMC. If out competitors (X-farm mining) can only source 100k chips per quarter while we can source 300k chips per quarter and our chips cost 60% of the cost to fab and run in their life-time than X-farms chips, then we can grow our farm above and beyond the capacity of X-farm to grow theirs.

http://www.taipeitimes.com/News/biz/archives/2013/12/13/2003578882

This also makes me want to invest in a foundry. Cheesy

Edit: I like your post.