3 ms·

It's not the value of bitcoins that stabilises but the difficulty of producing bitcoins (measured in average number of trial hashes per BTC). Given: D = diffi

by caf 15y ago

It's not the value of bitcoins that stabilises but the difficulty of producing bitcoins (measured in average number of trial hashes per BTC). Given: D = difficulty (HASH/BTC) E = marginal efficiency (HASH/kWH) V = value of bitcoins (USD/BTC) C = cost of electricity (USD/kWH) then in equlibrium, the difficulty will be equal such that the electricity you can purchase with a generated bitcoin will just balance out the electricity required to produce it (neglecting depreciation of equipment, time value of money etc..): D ~= (E * V) / C So this predicts that equilibrium difficulty will increase when marginal mining efficiency increases (ie. more power-efficient GPUs are available), when the value of bitcoins increases, and when the price of electricity decreases. (Note that there is no guarantee that the system will reach equilibrium - the relatively long time lag taken for the difficulty to adjust in the protocol may mean it doesn't). It is also notable that you can rearrange to solve for `V`, allowing you to determine an estimate of the underlying theoretical value of bitcoins based on the current difficulty level, the efficiency of the best GPU now available, and the price of electricity.

IsaacL 15y ago
Some people (eg, college students) have access to unmetered electricity, but we don't see such people dominating current mining efforts (AFAIK). I'm guessing this is because of the cost of hardware, so I think your model should also take into account the amortized cost of buying GPUs - though I don't know how significant that would be compared to the costs of electricity.
- caf 15y ago
  That is, of course, left as an exercise for the reader ;) (As long as the number of students with unmetered power partipating is small compared to the number of miners, it shouldn't significantly affect the analysis. It's the marginal cost faced by a new entrant that's of interest - a relatively small number of students mining simply adds a background base level of difficulty).
- fhars 15y ago
  It is the dominating cost for students with unmetered electricity, so it's probably not negligible. (Well, maybe it's only the second most important cost for them, after sleep loss due to fan noise in the dorm room, but that is difficult to put a monetary value on.)

by caf 15y ago

IsaacL 15y ago

Some people (eg, college students) have access to unmetered electricity, but we don't see such people dominating current mining efforts (AFAIK). I'm guessing this is because of the cost of hardware, so I think your model should also take into account the amortized cost of buying GPUs - though I don't know how significant that would be compared to the costs of electricity.

caf 15y ago
That is, of course, left as an exercise for the reader ;) (As long as the number of students with unmetered power partipating is small compared to the number of miners, it shouldn't significantly affect the analysis. It's the marginal cost faced by a new entrant that's of interest - a relatively small number of students mining simply adds a background base level of difficulty).
fhars 15y ago
It is the dominating cost for students with unmetered electricity, so it's probably not negligible. (Well, maybe it's only the second most important cost for them, after sleep loss due to fan noise in the dorm room, but that is difficult to put a monetary value on.)