Is Bitcoin viable, energy wise?

[phelix]

https://bitcointalk.org/index.php?topic=6459.0 concern for bitcoin and the environment
https://bitcointalk.org/index.php?topic=23360.0 Correlation between mining costs and Bitcoin value and ecological nightmare

I seem to remember I read something Satoshi said about this issue. Like he had to design the system in a way that would have a significant impact on world energy consumption. Can't find it now, though.

[cbeast]

https://bitcointalk.org/index.php?topic=6459.0 concern for bitcoin and the environment
https://bitcointalk.org/index.php?topic=23360.0 Correlation between mining costs and Bitcoin value and ecological nightmare

I seem to remeber I read something Satoshi said about this issue. Like he had to design the system in a way that would have a significant impact on world energy consumption. Can't find it now, though.

https://bitcointalk.org/index.php?topic=721.msg8114#msg8114

[mobodick]

https://bitcointalk.org/index.php?topic=6459.0 concern for bitcoin and the environment
https://bitcointalk.org/index.php?topic=23360.0 Correlation between mining costs and Bitcoin value and ecological nightmare

I seem to remember I read something Satoshi said about this issue. Like he had to design the system in a way that would have a significant impact on world energy consumption. Can't find it now, though.

Heya, thanks!

[myrkul]

What's your point?

As I wrote in the first post: few entities own and process all. The article shows it for a fact. Or do you have a difficulty reading it an understanding it?

No, but apparently you do. The article says nothing about processing. It says that someone made a large transaction back in November 2010, and many transactions since have been "related" (ie, used those coins). That doesn't even mean much, since of course those coins are going to be involved in a lot of transactions, they were a large portion of the economy at the time. Nearly every US dollar bill has cocaine on it. Does that mean everyone with a dollar in their pocket is a drug dealer?

Transactions are processed by miners. I could install a program on my PC, that, in a few minutes, could have me mining bitcoins using my graphics card. Just like that, I'm processing bitcoin transactions. That's decentralized.

The fast processing is for bitcoin payments. Not for getting money into or out of bitcoin.

Right, the latter is totally irrelevant. Why would you ever want to do that, and fast?
Oh yes, I am going to rush and buy 300 hard drives and 70 LCD screens when the price moves not to my liking?
Ever heard of Forex, where you can buy Euro and sell it in USD account in a microsecond?

And indeed, you can trade your Bitcoins for USD and back again in just seconds on Gox or any number of other exchanges - decentralized again - but you might have some problems getting the USD out of the exchange and into your pockets, though that's thanks to banking regulations. Back in the day, it was dead easy - if a little slow - bank transfers take so damn long.

Do you even know how much Visa charges? Or how bitcoin mining actually works?

What is great revelation about Bitcoin mining that makes you know what transaction fees they will be charging when it is the only source of income?

That they need fees. If they charge too much, there will be no transactions. No transactions means no transaction fees. They'll charge what they can get, and not a satoshi more. That's the way the Market works.

[mobodick]

https://bitcointalk.org/index.php?topic=6459.0 concern for bitcoin and the environment
https://bitcointalk.org/index.php?topic=23360.0 Correlation between mining costs and Bitcoin value and ecological nightmare

I seem to remeber I read something Satoshi said about this issue. Like he had to design the system in a way that would have a significant impact on world energy consumption. Can't find it now, though.

https://bitcointalk.org/index.php?topic=721.msg8114#msg8114

Satoshi makes a comparison to gold here.
One oversight he makes IMO is to forget that, unlike gold, once in circulation bitcoin still requires energy in various ways.

[johnyj]

Let's compare two cases:

1.
There are only 100 miners which have ASIC mining farm and each of their farm have a hashing speed of 10Th/s, the total network hashing speed is 1PH

2.
There are 1,000,000 miners which have GPU mining rig, average each of their mining rig have a hashing speed of 1Gh/s, the total network hashing speed is also 1PH

In such a case, the difficulty and hash rate is the same, the energy usage could be 100X difference, but is there any reason that the bitcoin value in the first case will be 100 times cheaper than the second case?

[firefop]

https://bitcointalk.org/index.php?topic=6459.0 concern for bitcoin and the environment
https://bitcointalk.org/index.php?topic=23360.0 Correlation between mining costs and Bitcoin value and ecological nightmare

I seem to remeber I read something Satoshi said about this issue. Like he had to design the system in a way that would have a significant impact on world energy consumption. Can't find it now, though.

https://bitcointalk.org/index.php?topic=721.msg8114#msg8114

Satoshi makes a comparison to gold here.
One oversight he makes IMO is to forget that, unlike gold, once in circulation bitcoin still requires energy in various ways.

OP - I think the premise of your original question is invalid.

The 'energy cost' of bitcoin network does not matter to the network and it never will. Precisely because miners are the ones providing that energy. Miners have already proven that the network hash rate will continue to increase (and thus the total energy expended in the act of mining) until it isn't profitable to increase it.

The goal of 'adoption' is to generate transactions fees that amount to greater than the block reward. There's no way for force an increase except by transaction size and this is completely fair to each user.

Assuming bitcoin is in general (global?) use the network hash rate will grow to any size that it needs to in order to process the volume of these transactions. Once this happens there won't be any disincentive to mining.

So energy is covered...


Now bandwidth, that's a different discussion... and may at some point become an issue... but I'd imagine we'll be able to develop some sort of supernode system or some bitcoin specific protocol to deal with this once it becomes a real concern.

~

Another thing to consider if we assume adoption... some of the big boys (aka other payment processing services) will eventually convert or at least support bitcoin. Once Amazon or Paypal take them directly they'll be contributing to the network as well (at least some fast nodes so they can send payments directly without using another service).

[myrkul]

Let's compare two cases:

1.
There are only 100 miners which have ASIC mining farm and each of their farm have a hashing speed of 10Th/s, the total network hashing speed is 1PH

2.
There are 1,000,000 miners which have GPU mining rig, average each of their mining rig have a hashing speed of 1Gh/s, the total network hashing speed is also 1PH

In such a case, the difficulty and hash rate is the same, the energy usage could be 100X difference, but is there any reason that the bitcoin value in the first case will be 100 times cheaper than the second case?

No. But you can make a reasonable assumption that the processing fees will be 100x lower.

[mobodick]

https://bitcointalk.org/index.php?topic=6459.0 concern for bitcoin and the environment
https://bitcointalk.org/index.php?topic=23360.0 Correlation between mining costs and Bitcoin value and ecological nightmare

I seem to remeber I read something Satoshi said about this issue. Like he had to design the system in a way that would have a significant impact on world energy consumption. Can't find it now, though.

https://bitcointalk.org/index.php?topic=721.msg8114#msg8114

Satoshi makes a comparison to gold here.
One oversight he makes IMO is to forget that, unlike gold, once in circulation bitcoin still requires energy in various ways.

OP - I think the premise of your original question is invalid.

The 'energy cost' of bitcoin network does not matter to the network and it never will.

But it might matter to society as a whole.

[mobodick]

Now bandwidth, that's a different discussion...

No, its actually a big part of what i'm interested in because all the networking consumes energy too.

[mobodick]

https://bitcointalk.org/index.php?topic=6459.0 concern for bitcoin and the environment
https://bitcointalk.org/index.php?topic=23360.0 Correlation between mining costs and Bitcoin value and ecological nightmare

I seem to remeber I read something Satoshi said about this issue. Like he had to design the system in a way that would have a significant impact on world energy consumption. Can't find it now, though.

https://bitcointalk.org/index.php?topic=721.msg8114#msg8114

Satoshi makes a comparison to gold here.
One oversight he makes IMO is to forget that, unlike gold, once in circulation bitcoin still requires energy in various ways.

Gold, once in circulation, still requires energy in various ways.

I guess you're right, but they do behave differently in this respect.
Bitcoin is supposed to become a currency and so i'd expect it to be transfered around a lot.
With bitcoin the energy expenditure does not depend on the amount transfered.
So even the smallest transactions will use up this energy.
I wonder if this will grow out of proportions and if not, where it will balance out.

But i can indeed see a more stable role for bitcoin as a sort of bulk transaction medium.
Then again, wouldn't that ultimately create bucketshop economies in front of it?

[firefop]

Now bandwidth, that's a different discussion...

No, its actually a big part of what i'm interested in because all the networking consumes energy too.

But it might matter to society as a whole.

I don't really know why you'd want to try and quantify this. Are you attempting to evaluate how 'green' bitcoin may or may not become? Bitcoin is such a small amount of traffic on the network relative to everything else that's going on there that I can't imagine it ever being an issue. Even if we imagined away all other currency and assumed all those trades were done using bitcoin instead... and scaled up the network accordingly.  it would still be so very minor compared to the bandwidth used by say facebook, google or netflix.

[mobodick]

Now bandwidth, that's a different discussion...

No, its actually a big part of what i'm interested in because all the networking consumes energy too.

But it might matter to society as a whole.

I don't really know why you'd want to try and quantify this. Are you attempting to evaluate how 'green' bitcoin may or may not become? Bitcoin is such a small amount of traffic on the network relative to everything else that's going on there that I can't imagine it ever being an issue. Even if we imagined away all other currency and assumed all those trades were done using bitcoin instead... and scaled up the network accordingly.  it would still be so very minor compared to the bandwidth used by say facebook, google or netflix.

The word 'green' is too political for me. I'd say i want to know the energy footprint of bitcoin.
Imagine a large part of the world is using bitcoin.
How much network traffic will it take for everybody to be up to date with everybody elses transactions?

[johnyj]

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  

True, you can do that.   But that case is the same as if you actually took your USD, and bought bitcoins at the exchange, instead of mining them. (And somebody else mined them).

I'm talking about the things that create imbalances on the exchange.  In your case, the exchange did not see the increased supply of bitcoins, but the cost of mining was still there, you paid it with your USD.

If a LONG TERM projection always indicate a rise in the exchange price due to limited supply, there will be less and less people sell bitcoin to cover their energy cost, and that trend itself will lift the bitcoin price as a result, can go on for decades

Another thing, your formula is based on that most of the people could always get their hands on the latest mining technology, thus the mining competetion will always be hard. But I foresee a future that only a handful IC companies have the access to the latest engergy efficient mining technology and many normal people with ASIC miners just can not cover their electricity cost, just like what happened after 10 years of california gold rush

This also partly answered OP's question, if the mining technology becomes more and more hidden and centralized, the actual energy cost will be less and less

[firefop]

The word 'green' is too political for me. I'd say i want to know the energy footprint of bitcoin.
Imagine a large part of the world is using bitcoin.
How much network traffic will it take for everybody to be up to date with everybody elses transactions?

According to http://blockchain.info/charts/blocks-size we're aproaching 6 gigabytes for the blockchain.
According to http://torrentfreak.com/bittorrent-and-netflix-dominate-americas-internet-traffic-111027/ Netflix is 34% of total...

I can't find numbers on exact transfer rate of the internet. But lets say ~1 exabyte per day (someone got a source to the real data?) so we're looking at netflix transfering 333,000 terabytes per day. This means that even if you had 55,500,000 bitcoiners downloading the entire blockchain every single day... you'd still be consuming less than netflix does on a daily basis.

For real numbers, number of tx * number of nodes * time = transfer rate of bitcoin network  ... feel free to include an estimation of non-current blocks being transfered to out of sync clients.

[mestar]

If a LONG TERM projection always indicate a rise in the exchange price due to limited supply, there will be less and less people sell bitcoin to cover their energy cost, and that trend itself will lift the bitcoin price as a result, can go on for decades

What long term projections?

I have yet to see one that is not wishful thinking, or a totally unrealistic "There is X total of something, lets divide that by 21,000,000 to see what we get."

And the mining incentives are built in into the system itself, and there is no need for random guesses.  Somebody calculated 2 years ago, that if the price goes to $20.000 you would need something like 60 power plants just for the energy for miners.

Another thing, your formula is based on that most of the people could always get their hands on the latest mining technology, thus the mining competetion will always be hard. But I foresee a future that only a handful IC companies have the access to the latest engergy efficient mining technology and many normal people with ASIC miners just can not cover their electricity cost, just like what happened after 10 years of california gold rush

Those with access to the technology would always increase the capacity until it is no longer profitable to do so.  The cost of mining per BTC will always be close to one BTC. 

This also partly answered OP's question, if the mining technology becomes more and more hidden and centralized, the actual energy cost will be less and less

How in the world would you suspend market forces in an open P2P network?

[mobodick]

The word 'green' is too political for me. I'd say i want to know the energy footprint of bitcoin.
Imagine a large part of the world is using bitcoin.
How much network traffic will it take for everybody to be up to date with everybody elses transactions?

According to http://blockchain.info/charts/blocks-size we're aproaching 6 gigabytes for the blockchain.
According to http://torrentfreak.com/bittorrent-and-netflix-dominate-americas-internet-traffic-111027/ Netflix is 34% of total...

I can't find numbers on exact transfer rate of the internet. But lets say ~1 exabyte per day (someone got a source to the real data?) so we're looking at netflix transfering 333,000 terabytes per day. This means that even if you had 55,500,000 bitcoiners downloading the entire blockchain every single day... you'd still be consuming less than netflix does on a daily basis.

For real numbers, number of tx * number of nodes * time = transfer rate of bitcoin network  ... feel free to include an estimation of non-current blocks being transfered to out of sync clients.

The difficult seems to be that when bitcoin grows the potential for transactions also grows. The more people there are in the bitcoin economy the more there is to transact in the bitcoin economy.
It's related to the problems you get with interprocess communication in networked processing.
I can see that the number of transactions will rise at some exponential number compared to a linera rise in number of nodes.
So the ammounts of transactions wil grow faster than the ammounts of nodes.
If bitcoin becomes anything like important in society then you will pay for your groceries with it, you'll pay for gas with it, you'll pay for the ticket to dysneyland with it. So there will be a lot more transactions per user to be done.
And all these transactions also add up to the block chain, so everyone will need to get the blockchain which includes all the transactions of everyone else in the world.
I can easliy predict that the 6G of chain we have now will be meaningless once bitcoin represents some substantial part of the economy.
The 6Gig will be a daily download for every user at best. It won't take long before you will need to get a gig per minute to be up to date.
We all think the blockchain has been growing pretty fast lately, right?
Well, right now the size is related to about 55000 transactions daily. You need to get up to date with 55000 transactions per day.
Now imagine 1/10th of the world uses bitcoin to pay for their usual things.
We get 1/10th of 7x10^9 = 7x10^8 people participating.
Let's take an average of 5 transactions per user per day. You need to buy gas, you get an icecream, you pay the rent, etc.
(i'm just using some ballpark figures here to see where we get)
This gives us 1.4x10^9 transactions daily. 1.400.000.000 transactions per day.
That means the blockchain will grow about 25000 times faster than it does now.

Lets make an estimate of how much the blockchain grows right now.

In the last 1.2 months the blockchain grew by about a gigabyte.
On average that is about 0.0269 gigabyte per day. This gives us about 27.5 MB per day.
Ok, so now let's multiply this by our projected change in magnitude.
We get 27.5 x 25000 = 672 gigabyte per day !!

So in the situation where 1/10 of the world population does 5 bitcoin transactions per day on average everyone will need to download 672 gigabytes of new blockchain per day to keep up. If you didn't update for 2 days you'll need to download over 1TB . If you didn't touch your wallet for a month you'll be pumping 20TB to get up to date. And all this to be able to use bitcoin as a currency.
After only a year of hoarding blockchain everyone would have about 240TB of blockchain stored locally.

I think this will be problematic from an energy point of view even before the practical problems of storage become too devastating.

[mobodick]

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  

True, you can do that.   But that case is the same as if you actually took your USD, and bought bitcoins at the exchange, instead of mining them. (And somebody else mined them).

I'm talking about the things that create imbalances on the exchange.  In your case, the exchange did not see the increased supply of bitcoins, but the cost of mining was still there, you paid it with your USD.

If a LONG TERM projection always indicate a rise in the exchange price due to limited supply, there will be less and less people sell bitcoin to cover their energy cost, and that trend itself will lift the bitcoin price as a result, can go on for decades

Another thing, your formula is based on that most of the people could always get their hands on the latest mining technology, thus the mining competetion will always be hard. But I foresee a future that only a handful IC companies have the access to the latest engergy efficient mining technology and many normal people with ASIC miners just can not cover their electricity cost, just like what happened after 10 years of california gold rush

This also partly answered OP's question, if the mining technology becomes more and more hidden and centralized, the actual energy cost will be less and less

But that will happen under circumstances where the gain of these new technologies is diminishing.
It will not be very profitable to invest in developing 0.0003% more efficient mining hardware and any power you think you can excert from your position will be fragile at best.
In the end you will have a lot of people competing over less and less possible gain.
The risks of not getting back your investments from producing new minig hardware will increase.

[johnyj]

What long term projections?

I have yet to see one that is not wishful thinking, or a totally unrealistic "There is X total of something, lets divide that by 21,000,000 to see what we get."

And the mining incentives are built in into the system itself, and there is no need for random guesses.  Somebody calculated 2 years ago, that if the price goes to $20.000 you would need something like 60 power plants just for the energy for miners.

I agree with your view of daily energy cost. It is a good way to estimate a worst case scenario where every one is having the same mining technology and just break even on energy cost

If each bitcoin worth $20,000, that is $72,000,000 per day, at $0.2/kwh, that is 360,000Mwh per day, about 15,000 Mw output, that is 3 big nuclear power plant

Those with access to the technology would always increase the capacity until it is no longer profitable to do so.  The cost of mining per BTC will always be close to one BTC.  

Why should they do that when no one is competing with them? Suppose the rest of the network is 6TH now, I could mine with 6TH with a daily profit of 1800 coins and a power consumption of 200 coins; or I could mine with 60TH, 3273 coins, but 10 times the power consumption of 2000 coins. Is there any motivation for the later? Well you need a chart to find out the optimum hashing power, but it is definitely not higher the better

[mestar]

Why should they do that when no one is competing with them? Suppose the rest of the network is 6TH now, I could mine with 6TH with a daily profit of 1800 coins and a power consumption of 200 coins; or I could mine with 60TH, 3273 coins, but 10 times the power consumption of 2000 coins. Is there any motivation for the later? Well you need a chart to find out the optimum hashing power, but it is definitely not higher the better

Once you have majority of the hashing power, there is no point in increasing it, because you are basically competing against yourself. (Also, if you are the only miner you could do it all on one CPU.)

However, no miner is in this position.  You could argue that all the miners could make a pact about the maximum capacity, but, there is simply no way to enforce this on a P2P network.

If a miner can increase his profit by adding more capacity, he will do so. And this arms race has a very easily predictable result.