There will be a sharper divide between people who are paying for power at scale and those who are using surplus power with no additional realized costs. Make friends with people who have long term contracts in data centers with unused power and empty cages, is my advice. 
Topic: Projected Minimum Cost per BTC over the next year - page 4. (Read 18793 times)
I predict the rise of the portable ultra efficient "my office or co-working space won't notice the power draw" miner There will be a sharper divide between people who are paying for power at scale and those who are using surplus power with no additional realized costs. Make friends with people who have long term contracts in data centers with unused power and empty cages, is my advice.
There will be a sharper divide between people who are paying for power at scale and those who are using surplus power with no additional realized costs. Make friends with people who have long term contracts in data centers with unused power and empty cages, is my advice.
it's also worth considering that many small timers might continue mining and hoarding at a production-price loss whereas mfger/megafarms that sell to cover payables simply would not.
perhaps difficulty increases will slow and the price will naturally follow your analysis, albeit at a more graduated pace? regardless, it will be very interesting to see what happens in the coming months
perhaps difficulty increases will slow and the price will naturally follow your analysis, albeit at a more graduated pace? regardless, it will be very interesting to see what happens in the coming months
I want to see some good analysis of the thermodynamics behind all this ...
... and if you cannot bring entropy, the second law and the irreversibility of information flow into it I'll probably dismiss it.
... and if you cannot bring entropy, the second law and the irreversibility of information flow into it I'll probably dismiss it.
Very interesting thread and analysis. It's crazy how much the cost to produce a bitcoin is increasing. Will definitely keep my eye on this thread.
If nothing gives (bitcoin price is below production cost) we know: Either at least some miners are economically irrational or we're making wrong assumptions about either the production cost or the price miners are able sell at.
Interesting times ahead
Interesting times ahead
I wouldn't assume some miners are economically irrational, I'd think it would reflect more on the nature of proponents of Bitcoin and how they value it.
Still this is another historic juncture.
"We Are Fuc*ed" I repeat We Are Fuc*ed"
Lol I see it as a win win.
"We Are Fuc*ed" I repeat We Are Fuc*ed"
So next September it will be 381 times harder to mine for BTC. Unless the price rises to $181,000 to make it equal to mining today. Something has to give.
Clearly the next step is to model amortized hardware costs. I would think that those will never become negligible. Perhaps it could modeled as some multiple of the electricity cost... for example, maybe the amortized hardware cost will always be twice the cost of power, so that the "apparent" cost of electricity is actually tripled.
Wow, your latest calculation about the price within different era's is really insightful!
I love this thread and all this number crunching. Awesome work !
This is all true at the current Bitcoin price. However, power consumption is a function of price.
There seems to be very little relationship between Bitcoin mining difficulty and price. In the last six months, difficulty has increased by a factor of 7, while price has declined a little. The cost of mining does not drive price. Price drives the level of mining activity. Total network power consumption is a function of price.
Mining difficulty and hash rate are a function of mining efficiency.
This is all true at the current Bitcoin price. However, power consumption is a function of price.
There seems to be very little relationship between Bitcoin mining difficulty and price. In the last six months, difficulty has increased by a factor of 7, while price has declined a little. The cost of mining does not drive price. Price drives the level of mining activity.
This is all true at the current Bitcoin price. However, power consumption is a function of price. As the price of Bitcoins goes up then network can "afford" to spend more on power.
Note that power consumption is a function of price not mining efficiency.
Mining efficiency only affects hash rate and difficultly - not power consumption.
For example, we cannot/do not want to get to $500,000 per BTC any time soon. Here is the math behind it:
https://bitcointalksearch.org/topic/estimating-the-energypower-consumption-of-the-bitcoin-network-694401
If BTC were to go to $500,000 in this era it would cause a catastrophic mining bubble:
$500,000 x 25 = $12,500,000 per block = $75,000,000 per hour
$75 million per hour would drive the mining to attempt to use 675 GW. This is about 30% of all the power generated on the planet.
So, in order to keep our power consumption under about 2% of world wide power production, we cannot/do not want the price to get to $500,000 before era 6, which is about 2033 or so.
Using my previously derived formula for the power consumption:
P = (6(50/2e) + f)(x)(1 - g)/c [kW]
where:
x = exchange rate [USD/BTC]
e = era [0..32] (we are currently in era 1)
f = average fees per hour [BTC/hour]
c = cost of energy [USD/kWh]
g = average gross profit margin [unitless ratio]
we can look at the power consumption in each era assuming a price of $500,000 per BTC.
In order to make it simple I will make the following assumptions:
x = $500,000 per BTC
f = fees per hour will keep the coinbase above 6 BTC/hour (1 BTC/block) in all eras
c = $0.10 per kWh
g = 0.1 miner gross profit margin
Note that power consumption is a function of price not mining efficiency.
Mining efficiency only affects hash rate and difficultly - not power consumption.
For example, we cannot/do not want to get to $500,000 per BTC any time soon. Here is the math behind it:
https://bitcointalksearch.org/topic/estimating-the-energypower-consumption-of-the-bitcoin-network-694401
If BTC were to go to $500,000 in this era it would cause a catastrophic mining bubble:
$500,000 x 25 = $12,500,000 per block = $75,000,000 per hour
$75 million per hour would drive the mining to attempt to use 675 GW. This is about 30% of all the power generated on the planet.
So, in order to keep our power consumption under about 2% of world wide power production, we cannot/do not want the price to get to $500,000 before era 6, which is about 2033 or so.
Using my previously derived formula for the power consumption:
P = (6(50/2e) + f)(x)(1 - g)/c [kW]
where:
x = exchange rate [USD/BTC]
e = era [0..32] (we are currently in era 1)
f = average fees per hour [BTC/hour]
c = cost of energy [USD/kWh]
g = average gross profit margin [unitless ratio]
we can look at the power consumption in each era assuming a price of $500,000 per BTC.
In order to make it simple I will make the following assumptions:
x = $500,000 per BTC
f = fees per hour will keep the coinbase above 6 BTC/hour (1 BTC/block) in all eras
c = $0.10 per kWh
g = 0.1 miner gross profit margin
Code:
Original target Subsidy Est Fees Power % of total world
Era starting year BTC/block BTC/hour GW power production
--- --------------- ----------- ---------- ----- ----------------
0 2009 50.00000000 0.00000000 1,350 58.41%
1 2013 25.00000000 0.00000000 675 29.20%
2 2017 12.50000000 0.00000000 337 14.60%
3 2021 6.25000000 0.00000000 169 7.30%
4 2025 3.12500000 0.00000000 84 3.65%
5 2029 1.56250000 0.00000000 42 1.83%
6 2033 0.78125000 1.31250000 27 1.17%
7 2037 0.39062500 3.65625000 27 1.17%
8 2041 0.19531250 4.82812500 27 1.17%
9 2045 0.09765625 5.41406250 27 1.17%
I find the following numbers most interesting. It shows a statistically significant, small, slowdown in build rate:
Code:
Average Average
Period Covered Per Adjustment
From To Adjust Length
--------- --------- ------- ----------
14-Mar-13 13-Mar-14 23.92% 11.38 days
29-Jun-13 29-Jun-14 23.45% 11.41 days
24-Aug-13 31-Aug-14 21.12% 11.63 days
Because it is becoming less profitable to mine. Or sell unprofitable machines to the public.
Or each additional machine added signifies a smaller and smaller increase overall.
yes going from 2ph to 4ph is no where close to 200ph to 202ph. one is 100 % diff the other is 1 % diff yet both are a 2ph in gear jump.
Also you need cheap power a 1 watt machine means 20 mega-watts for 20 ph.
So for the network to jump from 200ph to 220ph you need a new powerplant small one but a power plant none the less.
so to do a 20ph jump the plant below needs to be built (in the metaphorical sense for an analogy )
http://www.power-eng.com/articles/print/volume-111/issue-11/departments/managing-the-plant/repowering-a-small-coal-fired-power-plant.html
So down the road when and if we get to 400ph 10 percent is 40ph and means 2 of the above plants would need construction. If BTC does not crash and burn growth will slow. Or should I say power availability cannot keep up with the asic builders ability to build asics. What does it mean for BTC> wish I knew. But I am buying coins more then gear.
The total electricity used for Bitcoin mining is trivial when compared to worldwide power generation.
The total worldwide Bitcoin network only needs about 220 Megawatts for mining. A single typical large generator puts out 600 to 750 Megawatts and a typical power station would have several generators.
So I would not worry about electricity producers not being able to provide power for ASIC miners.
For example the Three Gorges dam complex in China has 32 generators each of 700 Megawatts. So the total worldwide Bitcoin mining network is equal to about 1% of the output of that one generating complex. Admittedly the Three Georges dam is a larger than average generating plant.
As you know China is the world's leading generator of electricity and has the most generating capacity (1,200,000 Megawatts - CIA fact sheet).
A comment on the link to the 20 megawatt plant referenced above. This is a very small plant and appears to have been mothballed since the above 2007 article. http://www.arpapower.org/lamar-repowering-project-plant-studies-2/
I find the following numbers most interesting. It shows a statistically significant, small, slowdown in build rate:
Code:
Average Average
Period Covered Per Adjustment
From To Adjust Length
--------- --------- ------- ----------
14-Mar-13 13-Mar-14 23.92% 11.38 days
29-Jun-13 29-Jun-14 23.45% 11.41 days
24-Aug-13 31-Aug-14 21.12% 11.63 days
Because it is becoming less profitable to mine. Or sell unprofitable machines to the public.
Or each additional machine added signifies a smaller and smaller increase overall.
yes going from 2ph to 4ph is no where close to 200ph to 202ph. one is 100 % diff the other is 1 % diff yet both are a 2ph in gear jump.
Also you need cheap power a 1 watt machine means 20 mega-watts for 20 ph.
So for the network to jump from 200ph to 220ph you need a new powerplant small one but a power plant none the less.
so to do a 20ph jump the plant below needs to be built (in the metaphorical sense for an analogy )
http://www.power-eng.com/articles/print/volume-111/issue-11/departments/managing-the-plant/repowering-a-small-coal-fired-power-plant.html
So down the road when and if we get to 400ph 10 percent is 40ph and means 2 of the above plants would need construction. If BTC does not crash and burn growth will slow. Or should I say power availability cannot keep up with the asic builders ability to build asics. What does it mean for BTC> wish I knew. But I am buying coins more then gear.
Jump to: