Block Size Scalability Issues

[Saxscraper]

     Let’s talk a little bit about the Bitcoin scaling issues. We will look at some projections for the future using estimates based on todays numbers. I realize that some of the numbers change quite a bit so it makes it a little difficult to calculate perfectly, but we can get some good generalizations. If I am missing something or if my logic is wrong or my math is wrong, please kindly help me understand, as this is meant for information purposes only and meant to provoke meaningful thought. Especially for Bitcoin core, who can not agree on how to scale, so, so far nothing has been implemented as far as block size is concerned. There is a lot of info to share So let’s begin…
    
     At current prices of $6,600 and with a block reward of 1800 coins a day (12.5 X 6 X 24) the current rewards paid out to the miners each day is $11,880,000. At some point in the future, around 2120 I think, the last bitcoin will be mined and the miners will have to live off of transaction fees alone. If that were to happen today, at todays rates, we would need to pay out the miners the roughly $11,880,000 with transaction fees alone. Something has to give. Either there will be less miners due to loss of profitability, or much higher network fees. At todays rate of $.572 average transaction fee (source https://bitinfocharts.com/comparison/bitcoin-transactionfees.html#3m ) we would need 20,769,230 transactions a day to pay the miners with the fees generated. We are currently just shy of 200,000 transactions a day. (source https://www.blockchain.com/en/charts) This means that we would need 103.8 times as many transactions a day to reach the 20,769,230 to make up the difference needed to sustain the miners. This comes out to 240.38 transactions per second! If you are familiar with the current block size limits you already know that the Bitcoin network can handle between 3.3-7 transactions per second. (source https://en.wikipedia.org/wiki/Bitcoin_scalability_problem) So, this means that the MINIMUM block size would need to be 35mb, not 1mb, not 2mb, not 8mb, but 35mb. No one is talking about a block size that large. No one. Again back to Bitcoin core. The block size is supposed to be scalable but they refuse to even double the size from current levels.
    
     These numbers listed above are conservative and if the price of Bitcoin goes up as high as some are speculating, they only get much worse. At a $50,000 Bitcoin price, and after the next Block reward halving (900 coins a day,) the amount of transactions would need to be 910.5 per second in order to pay the miners for their work if the block reward was gone. If we wanted to be at a point to where the network was growing at a rate that could keep up with miner payouts based on fees each time a block reward halving happens, then the network will need to grow to 910.5 transactions a second just to make up for the $45,000,000 lost in the block reward halving as well. So far we are not even close to that kind of adoption. Again, the network is currently capable of 3.3-7 transactions per second. Fast forward to 2120 when the last Bitcoin is mined and the numbers will have to be huge. What I am getting at is that in the end, if miners are going to live off of transactions fees alone, the current block size is only a tiny fraction of what will be required.  In the $50,000 Bitcoin price example shown above the block size would need to be 275.92 times as large to accommodate that many transactions. Bitcoin core will not even double the block size to 2 mb which would allow the Bitcoin network to reach a measly 6.6-14 transactions per second. What are they thinking? Some of the core members want to make no changes to Bitcoin and that means it will never see mainstream adoption. How could it? The numbers are simply impossible.

     One final thought: When the Bitcoin price rises mining becomes more profitable and the Hash rate will rise as more miners plug in to compete for the rewards. As prices fall, mining becomes less profitable and people will unplug their miners causing the hash rate to fall. It’s not the other way around. Just because the hash rate goes up it does NOT mean that the price of Bitcoin will also go up in order to support all the miners. It is possible to lose money mining Bitcoin.

This has been a brief summary of my thoughts The more we crunch these types of numbers the more obvious it becomes that the current situation on the Bitcoin network is a recipe for failure. The block size must be changed. We Better start asking the core guys really nice if they would like to go ahead and save the network, or are we doomed to live by their indecision forever? What are the motives to stay the same?      
    

[1714025891]

1714025891

[1714025891]

1714025891

[1714025891]

1714025891

[butka]

As far as I can tell, these calculations are correct if things stay as they are today.

At todays rate of $.572 average transaction fee (source https://bitinfocharts.com/comparison/bitcoin-transactionfees.html#3m ) we would need 20,769,230 transactions a day to pay the miners with the fees generated.

However, one may assume that the transaction fees are not going to stay at today's level. It may very well be that on-chain transacting on the bitcoin network won't be that cheap. Also, these calculations don't include off-chain scaling. We don't know now how big the impact of the lightning network will be and how it may affect the scalability problem.

[NeuroticFish]

The block size is supposed to be scaleable but they refuse to even double the size from current levels.

The history of last winter tells that the 1 MB block size may be better for miners than larger sizes because that can force a quite big rise in fees if the number of transactions grow enough.
So if you base your calculations on the assumptions that the number of transactions in 100 years will be at levels comparable the ones we have today and the price will also stay on this level, you are wrong. The price has to rise a lot (buzz words: deflationary coin, scarcity), the number of transactions (on chain and off chain should grow, maybe even exponentially)...

Then there's the difficulty. That may make my calculations wrong, but Bitcoin had in history moments when mining was not worth it anymore to certain categories. They stopped their gear and maybe the difficulty didn't rise that fast.  This could also happen in the future too if the price continues to drop.

[odolvlobo]

TLDR; You cannot assume that the value of the block reward in the future must be the same as it is today. You cannot assume that the maximum transaction rate depends directly on the maximum block size.

The security of the network depends directly on the value of the block reward. As such you must determine a minimum value of the block reward that will provide the minimum viable level of security. I don't know how to determine that value, but I'm sure that some smart person can figure it out. It depends on the value of the transactions, the maximum size of a block and the price of including a transaction in a block. I believe that the current value of the block reward exceeds the minimum necessary value by a large factor, so a large drop in the network hash rate going forward would have little impact on security.

It is possible to use Bitcoin without including every transaction in the blockchain. Already, a large portion of Bitcoin transfers are transacted off-chain through private channels provided by companies such as Coinbase and BitPay. In the future, the Lightning Network will provide another way to transact without going directly to the block chain, and other methods will probably be developed in the future. I don't think it would be crazy to predict that less than 1% of all bitcoin transfers will be done on-chain in the future.

[Foxpup]

Bitcoin core will not even double the block size to 2 mb

SegWit already quadrupled the maximum block size to 4 MB. Your argument is invalid.

[nc50lc]

Bitcoin core will not even double the block size to 2 mb

SegWit already quadrupled the maximum block size to 4 MB. Your argument is invalid.

Exactly!

And as an addition, we can also squeeze more inputs and outputs in a single transaction using SegWit addresses.
Quite anti-Mining? Caring too much about the price? (uhhh, I'm guilty too)
Bitcoin hasn't been paying miners to process transactions, they simply mine what's profitable.
This isn't the only SHA-256 coin out there, and the total hashrate won't affect the number of transactions per minute, just the network security.
Quitting miners: not much of a problem considering on how Bitcoin was designed.

The one who got a recipe of failure are Bitmain and ASIC Manufacturers, but they are currently bathing in profits for now.

The center of your attention based on your assessment should be: Bitcoin with low Network hashrate will be much more easier to attack with easy to access hardwares if/after most of the miners quit.

[ebliever]

    At current prices of $6,600 and with a block reward of 1800 coins a day (12.5 X 6 X 24) the current rewards paid out to the miners each day is $11,880,000. At some point in the future, around 2120 I think, the last bitcoin will be mined and the miners will have to live off of transaction fees alone. If that were to happen today, at todays rates, we would need to pay out the miners the roughly $11,880,000 with transaction fees alone. Something has to give. Either there will be less miners due to loss of profitability, or much higher network fees.    

It's not a problem for there to be less miners/hashrate, so long as it is distributed (decentralized). There's no reason to say that we have to pay miners some set amount or maintain parity with the current level of profitability.

[Kakmakr]

Ok, but did the Lightning Network not throw a fox into the chicken coop? Satoshi's idea was for the Block reward to decrease over time and for the tx's to increase, so that the miners rewards would eventually replace the Block reward? Now we are implementing a 2cnd layer solution that would effectively reduce micro payments from the Bitcoin Blockchain and this reduces the tx volume.

We are hoping the Lightning Network will solve the scaling limitations and that it will increase adoption, but what happens if this does not happen? We need to look to a future where the miners rewards needs to replace the Block rewards. 

 

[NeuroticFish]

Ok, but did the Lightning Network not throw a fox into the chicken coop? Satoshi's idea was for the Block reward to decrease over time and for the tx's to increase, so that the miners rewards would eventually replace the Block reward? Now we are implementing a 2cnd layer solution that would effectively reduce micro payments from the Bitcoin Blockchain and this reduces the tx volume.

We are hoping the Lightning Network will solve the scaling limitations and that it will increase adoption, but what happens if this does not happen? We need to look to a future where the miners rewards needs to replace the Block rewards. 

You are only partly right about the block rewards (and that means you are party wrong too).
Even yesterday, at a jump in price, people reacted heavily and the number of transactions grew a lot, also the tx fees started to rise. Since we aim to get to bigger numbers than VISA can handle, there's a good chance that average tx fees will be higher than the basic 1 sta/byte. You need patience though.
Then there's something else you've missed. The price of Bitcoin. If today the reward only from tx fees in a block is maybe 0.2 BTC, with high enough price of Bitcoin even this amount would do.
Bitcoin mining doesn't have to be for everyone(*), only the most efficient ones will keep up.

(*) If Bitcoin mining would have been for everybody, big changes would have been needed, see Monero.

[bob123]

Bitcoin core will not even double the block size to 2 mb

SegWit already quadrupled the maximum block size to 4 MB. Your argument is invalid.

SegWit introduced a maximum block weight of 4.000.000 units (compared to 1.000.000 bytes).

A 'typical' (non-segwit) byte in a transaction has a weight of 4 units.A witness byte weighs 1 unit.

So, theoretically if all transactions would be segwit we'd have 4 times more transactions inside the blocks than pre-segwit.
But the block size itself (for the non-witness part of the transactions still is at 1 MB.

[Atherus]

Bitcoin mining doesn't have to be for everyone(*), only the most efficient ones will keep up.

(*) If Bitcoin mining would have been for everybody, big changes would have been needed, see Monero.

This is a centralized approach to the issue though. Why not making available to anyone instead of excluding users from it?

[NeuroticFish]

Bitcoin mining doesn't have to be for everyone(*), only the most efficient ones will keep up.

(*) If Bitcoin mining would have been for everybody, big changes would have been needed, see Monero.

This is a centralized approach to the issue though. Why not making available to anyone instead of excluding users from it?

I agree that mining is today a business which also tends to be centralized. I may be also tempted to propose "mining for everybody". But unfortunately the things are more complicated.

Clearly, "mining for everybody" means a war against ASICs and Bitcoin may get out quite badly hit from such a war.
The coins that can be mined by average Joe tend to have very low price.
"Mining by everybody" will face accusations that main miners are botnets.
Changing the algo from time to time to stay ahead of ASIC manufacturers means forks which can be seen as signs against stability.

So unfortunately there's no really better alternative to the current way the mining goes. We can just hope that mining businesses can survive in different countries so the mining stays "somewhat decentralized" in the future too.

[Saxscraper]

Thanks everyone for contributing your thoughts in a respectful manner. It is a lot of information and I wasn't really sure where to start. I will look into the segwit block increase and do more research there.

However, people are pointing out issues about you can't use current rates for the fees. All the 2nd layer technology stuff like the lightening Network is only going to reduce fees drastically. Which means we will need even more transactions to pay as the block reward diminishes. It only validates what I am saying. If transaction fees drop to .02-.05 it will take many more to make the same amount of money for the miners than the current .57 transaction fee. 4mb block will not even be close.

These issues are still down the road a bit, but look what happened in December when the price shot up. Fees went through the roof, mempool back up grew, and business started dropping Bitcoin payments going into the new year. It is still early, but major adoption with the current setup would be fun to watch.

[odolvlobo]

... All the 2nd layer technology stuff like the lightning Network is only going to reduce fees drastically. ...

LN is not a perfect substitute for on-chain, so not all transactions will be done through LN.
Furthermore, increased adoption of Lightning Network will result in more on-chain transactions even if most transactions are off-chain because of the need to open and close channels. For example, if 10% of the world uses LN and opens/closes a channel once per year, that alone will generate 44 TPS, requiring (very roughly) 6 MB per block.
 

... Which means we will need even more transactions to pay as the block reward diminishes. ...

Again, the need for fees is determined entirely by the need for security (against a 51% attack). It is possible that the block reward will be more than enough to provide sufficient security even without a subsidy, but there is a real danger that it won't.

The issue is that security is not a major component in the price of a transaction. The primary factor is the size/weight, which is not related to its need for security. There needs to be a way to tie security to price. Some ways to increase the price of security might be to somehow make more valuable transactions bigger so that they cost more, or perhaps to include the value of transactions in determining the "longest" chain.

[Rath_]

... All the 2nd layer technology stuff like the lightning Network is only going to reduce fees drastically. ...

The Lightning Network makes Bitcoin micro-payments cost-effective. No one would pay a few cents or dollars for sending a few satoshis. Keep in mind that these transactions are nearly instant - it's something that won't be achieved on-chain. I guess that we won't bring zero confirmation transactions back.

Furthermore, increased adoption of Lightning Network will result in more on-chain transactions even if most transactions are off-chain because of the need to open and close channels. For example, if 10% of the world uses LN and opens/closes a channel once per year, that alone will generate 44 TPS, requiring (very roughly) 6 MB per block.

That's true. However, there are still many solutions which can increase the number of transactions per block (Schnorr signatures, MAST) without having to increase the block weight. Have you heard of channel factories? I have explained them here.

[Kakmakr]

Ok, but did the Lightning Network not throw a fox into the chicken coop? Satoshi's idea was for the Block reward to decrease over time and for the tx's to increase, so that the miners rewards would eventually replace the Block reward? Now we are implementing a 2cnd layer solution that would effectively reduce micro payments from the Bitcoin Blockchain and this reduces the tx volume.

We are hoping the Lightning Network will solve the scaling limitations and that it will increase adoption, but what happens if this does not happen? We need to look to a future where the miners rewards needs to replace the Block rewards. 

You are only partly right about the block rewards (and that means you are party wrong too).
Even yesterday, at a jump in price, people reacted heavily and the number of transactions grew a lot, also the tx fees started to rise. Since we aim to get to bigger numbers than VISA can handle, there's a good chance that average tx fees will be higher than the basic 1 sta/byte. You need patience though.
Then there's something else you've missed. The price of Bitcoin. If today the reward only from tx fees in a block is maybe 0.2 BTC, with high enough price of Bitcoin even this amount would do.
Bitcoin mining doesn't have to be for everyone(*), only the most efficient ones will keep up.

(*) If Bitcoin mining would have been for everybody, big changes would have been needed, see Monero.

This makes it even more risky, because we need two things to happen for this experiment to succeed. We need massive amounts of settlement tx's on-chain, when people open & close channels on the Lightning Network and the price of a bitcoin must grow exponentially for miners to be profitable. 

Fortunately for us, Bitcoin was developed to cover most of these challenges. If one thing goes down, another thing goes up and visa versa.  {Difficulty adjustment is one example}

[bob123]

We need massive amounts of settlement tx's on-chain, when people open & close channels on the Lightning Network and the price of a bitcoin must grow exponentially for miners to be profitable. 

With a big user base who is transacting daily on the LN, there will be enough on-chain transactions.
And if not, the fees for the on-chain transactions will be low enough for users to be attracted by on-chain tx's again.
I believe this will find an equilibrium.

The price does only have to grow exponentially if the miner do want to keep increasing the hashrate.
If mining gets unprofitable, the miner with the highest costs and lowest capital will stop mining and vice versa. This should also find an equilibrium.
 

[Samarkand]

...
Unfortunately it's impossible with current PoW algorithm/consensus method since people who have most efficient ASIC with cheapest electricity always win and in PoW, the winner takes all.
...

This may be true in theory, but in practice most miners are part of a mining
pool. Therefore they participate in the block reward every time that someone
from their pool "gets lucky" and manages to claim the block reward.

In the long-run this smoothes out the mining income and enables mining
on a more industrial scale. Of course there is still the possibility of mining
less than your expected share for a long time, but still taking part in a mining
pool reduces the variance inherent to a system that was designed as a
winner takes all system.

I wonder how Bitcoin would have turned out if we had never seen the emergence
of mining pools. The hashrate would probably be lower, because higher variance
would have made long-term planning and therefore industrial mining much more
difficult.

[cellard]

We need massive amounts of settlement tx's on-chain, when people open & close channels on the Lightning Network and the price of a bitcoin must grow exponentially for miners to be profitable. 

With a big user base who is transacting daily on the LN, there will be enough on-chain transactions.
And if not, the fees for the on-chain transactions will be low enough for users to be attracted by on-chain tx's again.
I believe this will find an equilibrium.

The price does only have to grow exponentially if the miner do want to keep increasing the hashrate.
If mining gets unprofitable, the miner with the highest costs and lowest capital will stop mining and vice versa. This should also find an equilibrium.
 

I don't know why but people often forget that on-chain transactions will continue happening every 10 minutes, without LN.

These that can afford on-chain transactions and see a point in doing them (most likely for bigger transactions that require it to be on-chain) will always find a good opportunity cost to use it and thus keep paying miners.

As usage goes up fees goes up, miners happy, hashrate goes up, users happy due safer blockchain. It's a nice snowball effect.

We all would like to be able to transact instant, cheaply on-chain without no sequences but apparently physics don't work that way, for now either pay the fee or use LN.

[Kakmakr]

We need massive amounts of settlement tx's on-chain, when people open & close channels on the Lightning Network and the price of a bitcoin must grow exponentially for miners to be profitable.  

With a big user base who is transacting daily on the LN, there will be enough on-chain transactions.
And if not, the fees for the on-chain transactions will be low enough for users to be attracted by on-chain tx's again.
I believe this will find an equilibrium.

The price does only have to grow exponentially if the miner do want to keep increasing the hashrate.
If mining gets unprofitable, the miner with the highest costs and lowest capital will stop mining and vice versa. This should also find an equilibrium.
 

I don't know why but people often forget that on-chain transactions will continue happening every 10 minutes, without LN.

These that can afford on-chain transactions and see a point in doing them (most likely for bigger transactions that require it to be on-chain) will always find a good opportunity cost to use it and thus keep paying miners.

As usage goes up fees goes up, miners happy, hashrate goes up, users happy due safer blockchain. It's a nice snowball effect.

We all would like to be able to transact instant, cheaply on-chain without no sequences but apparently physics don't work that way, for now either pay the fee or use LN.

Also, miners might start to supplement their income by hosting LN Nodes and getting fees from forwarding people's tx's. If they stop mining, then the Lightning Network will stop functioning and they will not get any fees from both the LN and the miners fees from their Bitcoin mining.

So in my opinion the Lightning Network actually supplement their income or it might just balance it out, when the Block reward decline.