Wednesday, December 11, 2013

Bitcoin Mining



I found a 2000GH/s mining rig for $6000 - the Black Arrow Prospero X-3, which is made in Shenzhen. Its chips are made with 28nm IC technology. Black Arrow will run the mining on it for 4 months for an additional $1500. Start-up date would be Feb 24.

This looked pretty good so I placed an order. But I didn't pay yet. I'll sit on this for a couple of days to ponder if I really want to risk my 9 BTC (today's value ~ $7500).

The gamble:


Its not clear what the mining difficulty levels will be over the next 6 months.  The rise rate in mining difficulty has slowed in the last 1-2 months.  Possible assignable cause: the big gains from 28nm have been realized and now they're in the past.  On the other hand, it could be the bend in the curve is nothing but a temporary slowing due to some other factor.

So I put together a spreadsheet model to calculate 4 month mining revenues for 3 different scenarios.  Each scenario has a different difficulty-monthly-increase-rate (highlighted in yellow above). The model predicted revenues from the mining are highlighted in rose.

The model says the red scenario (100% per month increase rate in difficulty, the dashed red line) is a loser -the mining rig costs 9 bitcoins and the revenues are less than that by 1.8 bitcoins.  For difficulty rise rates higher than 100% more than just 1.8 bitcoins will be lost.

The model says the green scenario, which assumes a 64% per month rate rise in difficulty as seen over the last 40 days (the green dotted line), will give a net return of nearly 5 bitcoins.

Hash rate, the total mining computing power, is also shown in the plot. The trend curves for difficulty and hash rate should parallel one another.  The plot shows they do - this is reassuring.  The slope from hash rate for the most recent 60 days (76% increase per month, the black dotted line) is possibly a better predictor of future mining difficulty. This is used in the black scenario - for this the model predicts a net return of 2.5 bitcoins. Not a great return, but OK.

There is a bit more return that can be had by mining another 2 months -  an additional 2 bitcoins for the 64% scenario, an additional 1.3 bitcoins for  the 76% scenario.


Bitcoin Denominations

With recent values of ~$1000 per bitcoin people have started speaking of "millibits".  A millibit is 1/1000 bitcoin. The current worth of a millibit is about one dollar.

The name bitcoin might be abbreviated to "bit".

If the bit were to rise to $1,000,000 (the value expected if all other currencies were to wither away and the bit was to fill in the emptiness), then a millibit would be worth O$1000 ("O" for old - dollars of today's dollar value), a microbit would be worth one Odollar, and a satoshi, which is 1/100 millibit, would be equivalent to an Openny.

want to buy ( or sell) some bitcoins?

Here's a service that's quite good.   It charges about a 1% comission.   It can be linked directly to your bank account.  coinbase


coinbase

2 comments:

  1. Has anyone computed the energy consumption of thw?

    ReplyDelete
  2. The computing power of all of the bitcoin mining operations as of 12/12/13 is 4,000,000 GH per second. The energy rate for a typical mining system is about 1 watt per GH per sec. So total energy consumption should be around 4 million watts. That's about the energy consumption of 4000 modern homes

    ReplyDelete

be sure to scroll down and hit the publish button when done writing