Bitcoin Forum

Anyone have any thoughts on what the 20nm will produce?

the newsletter says minimum of 3th/s

3TH 12.3BTC at the current price for Q1/Q2 2014

Q1/Q2 as a shipping date is a bad joke.. What does it means, that they could ship in January or in June? That's a 6 months time-frame. They should say in which month they ARE going to ship.

And then be off of their estimate by a single week and subject themselves to irate customers?

They are just giving themselves a good cushion in case there are problems which are very likely to occur.  And remember this is KNC... they under-promise and over-deliver.

*Over delivering late isn't over delivering. Its 'not getting sued'ivering.

my problem is i bought other systems for earlier delivery in that price range (january batch cointerra price was even less at $3/gh)

i am a (happy) customer of kncminer for their 1st jupiter batch and theyve performed excellently.  i want to be a repeat customer.   i was planning to be a repeat customer.   but i was hoping for a better $/gh - or an earlier delivery date (or both).... and i was expecting knc's next generation 20nm price per gh to be more competitive with the best 28nm out there.  i do realise that knc likes to under promise and over deliver... and i am expecting that again this time around... but if we have to rely on the 'over promising' margin to make the numbers work, its a difficult decision to make as we don't know what the margin will be.

meanwhile i ask myself how to justify buying 3TH systems from knc for $10k in march/april (q1/q2 they say) when I've already bought 2 TH systems from cointerra for $6k in jan..  i appreciate that those knc 3TH boxes could easily end up being 3.5 or 4 TH when delivered (and i hope they will be).  but that still only takes the price from $3.33/gh to $2.8/gh to $2.5/gh.  this isn't an easy decision.

If KNC use the same development model for this ASIC as their 28nm chip, I'm not sure that this isn't a bit of a show pony situation. It has often been pointed out already that the Bitfury 55nm chip is significantly more power efficient per GH/s than KNC's 28nm chip. It has been suggested that KNC didn't use a "full custom design", although I am unsure of the exact details of this. Whatever the reason, the Bitfury chip design was definitely better than KNC's rapidly developed effort (although the most common circuit board design for Bitfury is an inconsistent performer, some boards have exhibited all range of problems according to the forum chatter here).

Given their (again) short development time for what will be the leading processor geometry (20nm) in H1 of 2014, I'm not at all convinced at this stage that KNC aren't attempting the same development model again. The Cointerra or BlackArrow chips may well outperform these supposedly technically superior 20nm chips, just as the Bitfury chips did. And the other issues with the Bitfury implementation may not affect the newer 28nm offerings. It would be interesting to hear how the BlackArrow Bitfury boards are working out, they're the only other board/miner offering using the Bitfury chips that I am aware of.

I got the difficulty calculator out.
http://mining.thegenesisblock.com/a/22bd1742ff

put 3000GH
March delivery

it doesn't return ROI. Am I missing something?
Brian

20nm technology is not ground breakingly more efficient than 28nm. In fact, it brings more expenses to manufacturers and makes it difficult to compete with less expensive technologies.

For example, If the Bitfury + partners were a large company with lots of manufacturing capacity, they could with their current design put any 28nm or 20nm manufacturer out of business by pricing their miner units way below that what their hi-tech competitors can.

That said, I look towards seeing KnC's 20nm design. If it brings improvements to efficiency by design on the side of new technology, they have in their hands a product no other company is likely to pull off in years.

with a diff change per month of 100% that is too much I think....

we are at ~5PH now and diff will be 690Mio -> http://bitcoindifficulty.com/

just add the PH from KnC November Batch + HashFast and Cointerra till May. i think we will reach a diff around 2000Mio when KnC will deliver the neptunes. and at this diff 3TH will mine around 0,78BTC/day. lets think even worst, with a diff at 6000Mio you will earn 0,26BTC/day. i hope i have done no mistake. i appreciate every other opinion.

2100w hmmm I'm out.