Bitcoin reward for mining....some questions and hoping for your comments

[CliveK]

As mentioned in a previous post of mine, I am doing a thesis on Bitcoin, specifically the mining aspect to it, which is the background to my questions.  Please feel free to give your opinion/comment.  A different point of view is appreciated.

In the original paper by Satoshi Nakamoto, the writers acknowledge that an incentive is needed to support the network hence the pay out of newly minted coins (or transaction fees).  The reality is you need someone else’s computational power in order to keep the network up and running being a peer-to-peer system.  What the writer does not go into is the wasted CPU and/or electricity power of the miner who does not receive any minted coins for their contribution based on the all or nothing reward system.  With the current difficulty, needing specialised equipment to have a chance of success (ASIC hardware) and the increasing cost of power in most places, and the environmental impact of using more electricity (albeit that ASIC is more efficient than say CPU), do you think this inefficiency was ever considered?

With the writers referring to CPU, do you think they envisioned GPU or ASIC hardware being used to mine?  The difficulty has skyrocketed because of the increased hashrate due to popularity (BTC value) and more efficient hardware being used/massed produced.  I have just recently seen an article published by IEEE in 2004 entitled The Design of a High Speed ASIC Unit for the Hash Function SHA-256, so this technological concept predates Bitcoin by a good few years (2008/2009).

Next to Satoshi's paper, what do you recommend reading next?  Can anyone point to something more detailed on the reward system?

Thanks for taking the time to read this, your comments are welcome.

[jonnybravo0311]

As mentioned in a previous post of mine, I am doing a thesis on Bitcoin, specifically the mining aspect to it, which is the background to my questions.  Please feel free to give your opinion/comment.  A different point of view is appreciated.

In the original paper by Satoshi Nakamoto, the writers acknowledge that an incentive is needed to support the network hence the pay out of newly minted coins (or transaction fees).  The reality is you need someone else’s computational power in order to keep the network up and running being a peer-to-peer system.  What the writer does not go into is the wasted CPU and/or electricity power of the miner who does not receive any minted coins for their contribution based on the all or nothing reward system.  With the current difficulty, needing specialised equipment to have a chance of success (ASIC hardware) and the increasing cost of power in most places, and the environmental impact of using more electricity (albeit that ASIC is more efficient than say CPU), do you think this inefficiency was ever considered?

With the writers referring to CPU, do you think they envisioned GPU or ASIC hardware being used to mine?  The difficulty has skyrocketed because of the increased hashrate due to popularity (BTC value) and more efficient hardware being used/massed produced.  I have just recently seen an article published by IEEE in 2004 entitled The Design of a High Speed ASIC Unit for the Hash Function SHA-256, so this technological concept predates Bitcoin by a good few years (2008/2009).

Next to Satoshi's paper, what do you recommend reading next?  Can anyone point to something more detailed on the reward system?

Thanks for taking the time to read this, your comments are welcome.

As you correctly understand, the reward system is there to incentivize miners.  It is also there to get coins into circulation.  Remember, there exists a maximum number of coins that will ever be produced, and the generated coins for finding blocks is reduced as more blocks are solved.  For example, the initial reward for solving a block was 50BTC.  It is now 25BTC and will become 12.5BTC at block 420,000.  Eventually it will be the transaction fees that provide the incentive, and not the generated coins.

The "all or nothing" problem was solved with the advent of mining pools.  By pooling your hashing power with a group of people, the lot of you has a far better chance of solving a block than you do on your own.  Your reward is a portion of the generated coins and transaction fees, and is determined based upon certain methodologies (PPS, PPLNS, etc) employed by the pool.  For reference, take a look here: https://en.bitcoin.it/wiki/Comparison_of_mining_pools to learn the different payout mechanisms and their explanations.

[CliveK]

Hi jonnybravo0311,

Thanks for the reply, I appreciate your input.  I am aware of mining pools and have been testing two as part of my mining experiment.  Just another evolutionary step miners have had to undergo to stay relevant with the ever increasing difficulty, a hardy bunch.

I am currently doing a bit of research on the transaction fees.  All in all the fees seem pretty low, the blockchain seems to indicate it was around 12 BTC (https://blockchain.info/charts/transaction-fees) for call it 74k BTC in transactions when I last checked today, though the fees are not understood 100% by me yet...and how fees are currently charged since the fees are voluntary by the entity making the payment versus being charged by a 3rd party or your bank (https://en.bitcoin.it/wiki/Transaction_fees), though understand there is a reference implementation which will tackle this in future or is it currently being used now?

[SMB-2525]

I'm sure you read the transaction fee explanation.

My anecdote is when ordering my AntMiners, I doubled the default (and optional) recommended transaction fee. I had read about delays in transactions getting confirmed for AntMiners. I was very surprised at how rapidly my transaction was confirmed. I have no idea how the fee got split.

[jonnybravo0311]

Hi jonnybravo0311,

Thanks for the reply, I appreciate your input.  I am aware of mining pools and have been testing two as part of my mining experiment.  Just another evolutionary step miners have had to undergo to stay relevant with the ever increasing difficulty, a hardy bunch.

I am currently doing a bit of research on the transaction fees.  All in all the fees seem pretty low, the blockchain seems to indicate it was around 12 BTC (https://blockchain.info/charts/transaction-fees) for call it 74k BTC in transactions when I last checked today, though the fees are not understood 100% by me yet...and how fees are currently charged since the fees are voluntary by the entity making the payment versus being charged by a 3rd party or your bank (https://en.bitcoin.it/wiki/Transaction_fees), though understand there is a reference implementation which will tackle this in future or is it currently being used now?

Just like fees are voluntary by users, miners have the option to not pick up transactions.  This is why transactions that do not include fees take so much longer to confirm, if they ever do.  Miners simply don't pick them up because there's no incentive to do so.  In the future, as transaction fees play a larger role in miner incentives, far more miners will flat out refuse to process those transactions containing no fees.  Currently, it's not really an issue, because the reward of newly generated coins for finding a block far outweighs the transaction fees.

Another point to consider is the actual volume of transactions processed by the network.  At an average of 60,000 a day (which is about where we are currently), it's pretty minimal.  Even if every single transaction included the recommended fee of 0.0001BTC per kb of data, you're looking at 6BTC a day assuming 1kb transactions.  If you figure a block is solved every 10 minutes, that's 144 blocks a day, or an average of about 0.042BTC in transaction fees per block.  Not really awe inspiring.

Now, let's scale up a bit.  60,000 transactions a day is less than 1 transaction a second.  By comparison, Visa processes somewhere around an average of 2000 to 4000 transactions a second.  Ok, it might be a bit optimistic to assume the same volume, so let's say the network processes 100 transactions a second, which is about what PayPal does.  Using the same numbers as before (0.0001BTC per kb and 1kb transactions) we're looking at 864BTC a day in fees.  Same 144 blocks solved, and we've got 6BTC in fees per block.