Is Bitcoin viable, energy wise?

[mobodick]

I'd like to start a discussion about the viability of bitcoin in the future on basis of energy.

Bitcoin is set on a path that requires increasing amounts of computation to function.
Moreover, it requires the internet (or some other high speed network) to operate.
All this computing and communicating costs energy.
Growth of the bitcoin network implies more nodes communicating with more nodes. The posibility space for transactions will grow faster than the number of nodes.

I wonder if our global energy production can keep up with this requirements to get bitcoin settled as a solid world wide coin.
For one, i don't expect the hashing algorithm to become much more efficient. It will become a limit at some point so in the near future you can only increase hash rate by increasing energy consumption by the same amount.

The obvious question is, how will the energy requirement of running the bitcoin network (including the communications over internet) develop over time given a certain growth rate of transactions?

So what would be the best way to judge the energy consumption of bitcoin?
Knowing the energetic cost of a single transaction seems like a good starting point, but it would have to be based on some mean which i don't have.
But we propably will need to separate some parts of the process. There is the initiation of the transaction that requires that you have downloaded (some part of) the blockchain, there is the processing of the transaction by the network and there is the receiving and reading of the blockchain at the other side of the transaction.
The cost of downloading the blockchain is also somewhat separate.

I will try looking for some concrete figures but any help would be appreciated.

[mobodick]

Are armored trucks, viable energy wise?

No, not on a large scale. Even SUV's are not viable on any real scale.

So one of my concerns is that bitcoin will become energy bound because that would put a pressure on the price of transacting a small ammount. Transfering small amounts is required to make bitcoin a currency. It costs you very little energy to carry cash or a bank card. Bitcoin requires multiple computers and a whole lot of actual computation to even do one transaction. It is much more expensive energy wise than any other form of common currency.

[Le Happy Merchant]

A Dyson sphere or two would kick the can down the road a bit I suppose.

Please see: Liquid fuel Thorium reactors.

[gopher]

Very original thought...

I think in the near future (10-50 years) ASICs at 10nm-1nm will be energy efficient enough.

Also, I do not think that Bitcoin requires infinitely increasing number of nodes to function, only needs large number of nodes with sufficient hashing power to prevent someone from mounting a 51% attack.

The currently established model, having few large pools containing majority of the hashing power, I think will be an adequate match to the network's growth challenge to quite a degree.

[mobodick]

A Dyson sphere or two would kick the can down the road a bit I suppose.

But if seriously, I guess free market will sort out how much energy should humankind spend on monetary system. There is no need to go completely insane on this. We so far, for comparison, managed nuclear paritet without having a nuke in every bedroom.

No, the free market will use up any available energy because energy is the stuff that gets things done.
We, as a human species, already use much more energy than is provided by the sun and bitcoin makes an increasing contribution to this energy consumption.
Nuclear energy is basically out of the question for some time. The old reactor designs are breaking apart, we produce more nuclear waste than we can reasonably let cool down and new designs are expensive because they do not operate on the byproducts of keeping a nuclear arsenal.

Maybe i could rephrase the question.

How much can bitcoin grow before becoming too heavy for the free market to carry?

[JimiQ84]

A Dyson sphere or two would kick the can down the road a bit I suppose.

But if seriously, I guess free market will sort out how much energy should humankind spend on monetary system. There is no need to go completely insane on this. We so far, for comparison, managed nuclear paritet without having a nuke in every bedroom.

No, the free market will use up any available energy because energy is the stuff that gets things done.
We, as a human species, already use much more energy than is provided by the sun and bitcoin makes an increasing contribution to this energy consumption.
Nuclear energy is basically out of the question for some time. The old reactor designs are breaking apart, we produce more nuclear waste than we can reasonably let cool down and new designs are expensive because they do not operate on the byproducts of keeping a nuclear arsenal.

Maybe i could rephrase the question.

How much can bitcoin grow before becoming too heavy for the free market to carry?

Beauty of free market is that it can carry anything.

[mobodick]

Very original thought...

I think in the near future (10-50 years) ASICs at 10nm-1nm will be energy efficient enough.

Also, I do not think that Bitcoin requires infinitely increasing number of nodes to function, only needs large number of nodes with sufficient hashing power to prevent someone from mounting a 51% attack.

The currently established model, having few large pools containing majority of the hashing power, I think will be an adequate match to the network's growth challenge to quite a degree.

I don't see how quantum mechanical effects could be suppressed to make 1nm technology viable. Expect it to stop on a few nanometers.

Bitcoin will need to keep growing its computational power because the bigger it gets the more value it has to succeed in a 51% attack. So the pressure to break the network will increase with its market cap.
It is effectively an arms race.
So while it is true that we only need enough power to prevent an attack, the attack severity will also increase.

But i was interested in the energetic picture.
Not sure how the model of pool operators helps clearing that up.

[Nancarrow]

A Dyson sphere or two would kick the can down the road a bit I suppose.

Please see: Liquid fuel Thorium reactors.

You know, you could just google 'Dyson sphere'. 

[Nancarrow]

We, as a human species, already use much more energy than is provided by the sun...

You are making this mathematics and physics tutor *very* annoyed.

[Ekaros]

Very original thought...

I think in the near future (10-50 years) ASICs at 10nm-1nm will be energy efficient enough.

Also, I do not think that Bitcoin requires infinitely increasing number of nodes to function, only needs large number of nodes with sufficient hashing power to prevent someone from mounting a 51% attack.

The currently established model, having few large pools containing majority of the hashing power, I think will be an adequate match to the network's growth challenge to quite a degree.

I don't see how quantum mechanical effects could be suppressed to make 1nm technology viable. Expect it to stop on a few nanometers.

Bitcoin will need to keep growing its computational power because the bigger it gets the more value it has to succeed in a 51% attack. So the pressure to break the network will increase with its market cap.
It is effectively an arms race.
So while it is true that we only need enough power to prevent an attack, the attack severity will also increase.

But i was interested in the energetic picture.
Not sure how the model of pool operators helps clearing that up.

I don't see mining as a real issue. Probably with very wide adoption one party can't easily reach 51% and most of mining is done, by financial institutions to keep their fees down.

Real issue in my mind is the cost on network from traffic viewpoint. If individuals can use their own clients and everything is highly connected the needed throughput could be quite large. Though likely even this isn't an issue. 10MB blocks is around 1MB/s which is 1,3Mb/s. For 70 000 000 clients(1 client for every 100 people) this is 91.4Tb/s. Which is quite large amount of traffic, on other hand it could be done by using such techniques as multicast... So not bad for future...

[mestar]

I'd like to start a discussion about the viability of bitcoin in the future on basis of energy.

Bitcoin is set on a path that requires increasing amounts of computation to function.
Moreover, it requires the internet (or some other high speed network) to operate.
All this computing and communicating costs energy.
Growth of the bitcoin network implies more nodes communicating with more nodes. The posibility space for transactions will grow faster than the number of nodes.

Energy spent on mining will dwarf all other costs.  Also, this energy is easy to predict, because energy spent on mining is a simple function of Bitcoin price and total block award.  This is because mining has very low barrier of entry, and this ensures that as long as it is profitable, new miners will join, until the most inefficient miners are at the break-even point.  And the most profitable miners will try to expand their operation.


This makes it that each block costs somewhere around  (0.5 * block award * price) to (0.75 * block award * price).  And since awards are paid in Bitcoins and electricity is paid in local currencies, this makes the situation that all mining costs hit the exchanges every day, and push the price down.


This can currently be estimated to be $150.000 to $250.000 each day (using price of $90), and this amount of fresh money must enter the exchanges every day, or the price will go down. 


At this point we have a temporary delay in the "mining difficulty follows price" process , because of the change in technology and delays in deliveries of ASIC miners, so mining is very profitable at this point, and this reduces the selling pressure on exchanges.  This, plus block award halving is why we have this current Bitcoin bubble.

[mestar]

Also, it can be estimated that the total energy cost of mining all 21 million Bitcoins will be about 10 million to 15 million Bitcoins.

[Elwar]

A Dyson sphere or two would kick the can down the road a bit I suppose.

Please see: Liquid fuel Thorium reactors.

You know, you could just google 'Dyson sphere'. 

I am going to get myself a Dyson sphere to run my mining rig. Then I will have the best mining rig on the plant.

[myrkul]

A Dyson sphere or two would kick the can down the road a bit I suppose.

Please see: Liquid fuel Thorium reactors.

You know, you could just google 'Dyson sphere'. 

I am going to get myself a Dyson sphere to run my mining rig. Then I will have the best mining rig on the plant.

You'll have the best planet on the planet.

[mobodick]

I'd like to start a discussion about the viability of bitcoin in the future on basis of energy.

Bitcoin is set on a path that requires increasing amounts of computation to function.
Moreover, it requires the internet (or some other high speed network) to operate.
All this computing and communicating costs energy.
Growth of the bitcoin network implies more nodes communicating with more nodes. The posibility space for transactions will grow faster than the number of nodes.

Energy spent on mining will dwarf all other costs.  Also, this energy is easy to predict, because energy spent on mining is a simple function of Bitcoin price and total block award.  This is because mining has very low barrier of entry, and this ensures that as long as it is profitable, new miners will join, until the most inefficient miners are at the break-even point.  And the most profitable miners will try to expand their operation.


This makes it that each block costs somewhere around  (0.5 * block award * price) to (0.75 * block award * price).  And since awards are paid in Bitcoins and electricity is paid in local currencies, this makes the situation that all mining costs hit the exchanges every day, and push the price down.


This can currently be estimated to be $150.000 to $250.000 each day (using price of $90), and this amount of fresh money must enter the exchanges every day, or the price will go down. 


At this point we have a temporary delay in the "mining difficulty follows price" process , because of the change in technology and delays in deliveries of ASIC miners, so mining is very profitable at this point, and this reduces the selling pressure on exchanges.  This, plus block award halving is why we have this current Bitcoin bubble.

I don't see how energy (Joules, Watt) is a function of bitcoin price ($) and block reward ($?).
I'm looking for something like the mean minimum energy per hash, in the future.
Then all i need to figure out is how much hashing needs to be done to sustain a certain size network and i would get a number for the miners.

[mestar]

Copied from "Speculation"

Read your linked post. I think you're disregarding one-time cost of the asic miner unit. With asic the one-time cost is the large factor, energy cost less so. At least for now.

So I'm guessing energy used for mining should be considerably less than $2 to $3 million when rate is at $1000/BTC.

Difficulty grows with total hash rate, it doesn't care about power efficiency per hash.

Once you have the hardware, it makes economic sense to keep mining as long as your costs per one Bitcoin mined are less then one Bitcoin.

And if your cost is, for example 0.3 coins per coin, it makes sense to expand (in a sense of buying new hardware).  And buying new hardware will increase difficulty up to the point, perhaps at 0.5, perhaps higher, where you stop expanding.

Those two incentives create a situation where you will always have some miners that are close to zero profit, and not very many of those that are spending 0.1 BTC to mine 1 BTC.

We can never know the true efficiency distribution of all miners, but if you take a straight line from zero to one, on a chart where you display efficiency of all miners, sorted by efficiency, you would get total surface of 0.5,  and if you start from an assumption that the most efficient miner is at 0.5, you would get total of 0.75 under the curve.  Thus my assumption for total miner efficiency of 0.5 to 0.75.   In practice, it could be higher,  but if it is lower, it would be a temporary situation.

[mobodick]

We, as a human species, already use much more energy than is provided by the sun...

You are making this mathematics and physics tutor *very* annoyed.

LOL., yeah, sorry about that one. I was thinking about the oil economy and how much energy is captured as organic material that will some day produce fossil fuels.

[mobodick]

Copied from "Speculation"

Read your linked post. I think you're disregarding one-time cost of the asic miner unit. With asic the one-time cost is the large factor, energy cost less so. At least for now.

So I'm guessing energy used for mining should be considerably less than $2 to $3 million when rate is at $1000/BTC.

Difficulty grows with total hash rate, it doesn't care about power efficiency per hash.

Once you have the hardware, it makes economic sense to keep mining as long as your costs per one Bitcoin mined are less then one Bitcoin.

And if your cost is, for example 0.3 coins per coin, it makes sense to expand (in a sense of buying new hardware).  And buying new hardware will increase difficulty up to the point, perhaps at 0.5, perhaps higher, where you stop expanding.

Those two incentives create a situation where you will always have some miners that are close to zero profit, and not very many of those that are spending 0.1 BTC to mine 1 BTC.

We can never know the true efficiency distribution of all miners, but if you take a straight line from zero to one, on a chart where you display efficiency of all miners, sorted by efficiency, you would get total surface of 0.5,  and if you start from an assumption that the most efficient miner is at 0.5, you would get total of 0.75 under the curve.  Thus my assumption for total miner efficiency of 0.5 to 0.75.   In practice, it could be higher,  but if it is lower, it would be a temporary situation.

Ok, but what total hashrate is needed to sustain a network of, say, 1/10 of the world economy/

[johnyj]

This makes it that each block costs somewhere around  (0.5 * block award * price) to (0.75 * block award * price).  And since awards are paid in Bitcoins and electricity is paid in local currencies, this makes the situation that all mining costs hit the exchanges every day, and push the price down.

Why should I pay the electricity with mined bitcoin? I can use fiat to pay the electricity and save the coins, if I believe bitcoin price will rise in long term due to scarcity

Imagine that last year my electricity cost is 2000 USD, and I mined 200 coins, I paid electricity using USD. Now I only need to sell 22 bitcoin to pay back that amount of dollar. If I paid using bitcoin as you suggested, I would have paid 100-150 bitcoins for my electricity cost  

[mestar]

I don't see how energy (Joules, Watt) is a function of bitcoin price ($) and block reward ($?).
I'm looking for something like the mean minimum energy per hash, in the future.
Then all i need to figure out is how much hashing needs to be done to sustain a certain size network and i would get a number for the miners.

By, "energy spent", I meant "the cost of energy spent".   Now I see that I indeed wrote it wrong, sorry about that.


Miners don't care about how much hashing needs to be done, they only care about their profit.

And if you have a new generation of miners, and your cost is 0.01 BTC per BTC mined, you and other miners have a strong incentive to expand operations.

We are currently in the process of moving to ASICs, and this is going kind of slow, and that is one of the two reasons of the current bubble (the other being halving the award to 25.)