Topic: Channeling the ASIC manufacturers (Read 2981 times)

2 points
-The point at which miners go negative on power is not necessarily when the miner outlives its useful life.  Miners, be they small home miners or large industrial farms still are paying for the space the miner occupies, the energy needed to cool the mining space, and the acoustical noise of the miner.  Some home miners could benefit from the heat during winter and industrial farms don't care about noise - but these are factors that would cause a miner to be trashed earlier or later.
-While SP30s in your example cost us X, they cost the manufacturer some percentage of X.  If the markup and shipping was 50% of the cost, Spoondoolies would have only 1/2 the production cost of each miner.  Then the factors noted above apply

So if all the manufacturers were charging 2 to 3 times what it cost to make the miner, they can still easily profit from mining directly on their own farms.

No disagreements. I wonder what they do next? What is potential payback to developing the next  generation of ASICs?

At 15% and 12 days per bump, $0.10/KWH and $600BTC, the SP30s go negative on power in 204 days after barely paying for themselves. IMHO they need the value of BTC to go up or the rate of difficulty increases to go down in order to invest in a new generation of ASICs.

Any ASIC manufacturer that is currently alive has already sucked money out of miners and they're sitting on tons of cash and possibly viable efficient chips.  Why would they bother wasting money and shipping miners to consumers when they can just build a farm and mine their own BTC.  They already raised their capitol with their little pre-order games - now they just need to figure out if they think BTC will stay alive.

I am trying to get into the heads of the ASIC manufacturer/miners and draw some conclusions about where they think mining is going. To that end I built a tool where I could model various scenarios. This run is focused on power consumption verses when the device could go into production. Most of the relationships are linear, so you can insert your own assumptions for BTC$, efficiency, power cost and days per difficulty bump.

The first column is the number of difficulty bumps from the start (which is current difficulty). At 12 days per bimp, this is 48, 96 and 144 days. It reflects when new equipment might go into production.
Second column is difficulty
Third column is always one tera-hash
Forth column is the Kilowatts for 1 tera-hash
Fifth column is the BTC conversion. This is used to compute the power crossover point.
Sixth column is an efficiency factor. It assumes the one terahash is not always available.
Sixth column is the cost per kilowatt hour of electricity
Seventh column is the increase in difficulty with each bump. It has been running 16% in 2014
Eighth column is the number of days between bumps. This has been running 12.1 in 2014
Last column is the dependent variable - it is the number of days until the miner goes negative on power cost.

There are four groups of starting difficulty under the 15% & 12 day assumption: today, 48 days from now, 96 days from now and 144 days from now.

Interpreting the table:
If you put a one tera-hash miner in production today that consumed one kilowatt per tera-hash, it would remain profitable for 153 days.

If you put a one tera-hash miner in production in 48 days that consumed one half kilowatt per tera-hash, it would remain profitable for 165 days.

If you put a one tera-hash miner in production in 144 days that consumed one quarter kilowatt per tera-hash, it would remain profitable for 128 days.