How would you design/ calculate the setup ( cpu power) for national CC?

[kaffeeladen]

I´m writing my bachelor´s thesis with the working Title:
 Could Germany reduce its CO2 Balance by replacing coins and banknotes with Bitcoin?

therefore I want to ask:

How would you design/ calculate the setup ( cpu power) for national cryptocurrency- when 125´000´000 transaktion per day have to be executed? ( 11 nodes/ miners and alternatively with 30000 nodes/ miners)?

thank your very for your guidance and help I already recieved.





edited old post_ i was young and did not know anything:  i still don´t anything know but i´m older now.

(1. calculate how much energy only Germany would use when utilizing the bitcoin network per year with its 600 million transactions

2. what would take to set up a hypthetical bitcoin network  only for germany in germany, so that i could derive the CO2 emissions based on germanys electricity production infrastructure.

how can i calculate the needed hashpower ?

I´m sure that didn´t consider all aspects of the question.


any kind of hints and help is highly appreciated!

[aplistir]

I want to
1. calculate how much energy only Germany would use when utilizing the bitcoin network per year with its 600 million transactions

2. what would take to set up a hypthetical bitcoin network  only for germany in germany, so that i could derive the CO2 emissions based on germanys electricity production infrastructure.

how can i calculate the needed hashpower ?

You can't calculate the exact hashpower that would be needed.
Blockchain adjusts the required hashrate about once in every 2 weeks. That means that regardless of the used hashrate about 1 new block is found in 10 minutes (on average).
You can use as much hashpower as you want and still on average only 1 new block will be found every 10 minutes

If Germany would set up its own blockchain, then the hashrate could be much smaller. Especially if it is not made too interesting to existing mining farms. With less competition the difficulty would be easier.

Making transactions in bitcoin network do not really need much energy. What needs energy is competing with other big mining farms. If all of them would slow down their hashrate, we would still be able to make as many transactions, but the energy "requirements" of blockchain would be much smaller.

Competition takes all the energy not the blockchain itself.
 

[kaffeeladen]

I want to
1. calculate how much energy only Germany would use when utilizing the bitcoin network per year with its 600 million transactions

2. what would take to set up a hypthetical bitcoin network  only for germany in germany, so that i could derive the CO2 emissions based on germanys electricity production infrastructure.

how can i calculate the needed hashpower ?

You can't calculate the exact hashpower that would be needed.
Blockchain adjusts the required hashrate about once in every 2 weeks. That means that regardless of the used hashrate about 1 new block is found in 10 minutes (on average).
You can use as much hashpower as you want and still on average only 1 new block will be found every 10 minutes

If Germany would set up its own blockchain, then the hashrate could be much smaller. Especially if it is not made too interesting to existing mining farms. With less competition the difficulty would be easier.

Making transactions in bitcoin network do not really need much energy. What needs energy is competing with other big mining farms. If all of them would slow down their hashrate, we would still be able to make as many transactions, but the energy "requirements" of blockchain would be much smaller.

Competition takes all the energy not the blockchain itself.
 

Would it be sensible that assuming setting up only one mining center with redundant backup ( geographicaly located in a different part of germany) so that no competition must taken into account?

[ranochigo]

The validating of transactions uses relatively little resources. If the difficulty is the same, the theoractical resources used for generating a block is the same, even if one block has a thousand transaction and the other has one. The difficulty sets a requirement for the miner to meet. The higher the difficulty, the longer it would take, if the miner has no change in hashrate and ceteris paribus.

Mining is basically just the miner hashing the block header which contains the merkle root. The merkle root is the hash of the transactions in the block. Your ASIC does not validate transactions. In fact, if you run a Bitcoin node, you are also validating transactions the same way as a miner. I doubt you can accurately measure the energy consumption of miners in Germany at all. There is no indication of the location of where the mining farms or miners is located.

[nc50lc]

Would it be sensible that assuming setting up only one mining center with redundant backup ( geographicaly located in a different part of germany) so that no competition must taken into account?

If that's the case, a single Pentium4 PC can process 600 Million transactions per year; in exaggeration, set the difficulty to 1 and you can mine with an abacus.
The problem with that setup is it only requires a single person or group to manage the (centralized) miners which must be trusted by the whole network.

Of course, we will benefit from no/low transaction fee, low network power consumption, anything you can think of.
But with that, if a single corrupt person or an attacker got his hands on the miners, the whole network will be the victim.
Everyone can't even trust the most trusted person in the world: Dave.

Bitcoin was designed to be "trustless" or decentralized, the ledger must be "copied" exactly and verified by each of the nodes, new blocks must be verified by them, and "mining" must be done in a decentralized manner too.
All in order to be "trustless" without the requirement of a middle man or a centralized operator.

[buwaytress]

Best answer to refer to is nc50lc. @OP Perhaps what you're asking here is an imagined scenario where a single entity in Germany sets up a blockchain and is the only one securing the network. In which case, as nc50lc points out, even an abacus would be enough to solve the transactions so computing power is not even necessary - the weakest computer available today would effortlessly solve all the transactions necessary.

You will need to understand, however, that it's pointless, inefficient to have just one miner securing your centralised, private blockchain. There's a reason Bitcoin is so secure today, and getting more secure as more mining power enters the network.

But for the sake of calculating anyway, it should be pretty straightforward in your case. How much power you'd consume would be based on your block times (which would determine how long you'd need to confirm), and how many txs you could fit into your block (that also depends on your tx size then). But if you're already set at 600m txs per year... the current Bitcoin protocol couldn't process 600m a year.

At best, Bitcoin can process about 7 txs a second (though I think even at peak efficiency it only managed around 4). That's only about 220m txs a year. Your blockchain would need to make blocks about 3 times faster.

[HeRetiK]

1. calculate how much energy only Germany would use when utilizing the bitcoin network per year with its 600 million transactions

The transaction amount is irrelevant to the energy expended for securing a PoW blockchain as it has nothing to do with the hashrate.

Ignoring Moore's Law, the energy cost of a blockchain depends on the hashrate which depends on how much miners are willing to pay for electricity. How much miners are willing to pay for electricity, depends on how much the blockchain pays their miners. How much the blockchain pays their miners, has little to do with the transaction throughput, assuming that transaction fees are negligible and mining subsidy ie. block reward is significant.

For example: If a blockchain pays its miners EUR 10,000,000,- a year, and EUR 100,000,- is an acceptable profit, miners will spend EUR 9,900,000,- a year on electricity. If a blockchain pays its miners only EUR 1,000,- a year, and EUR 100,- is an acceptable profit, miners will spend only EUR 900,- a year on electricity. Keep note though, that the latter is much cheaper to attack, thus more vulnerable.

The example above is over-simplified of course and ignores hardware and infrastructure costs, as well as the volatility of cryptocurrencies. I do hope it helps exemplify why transaction throughput is a bad starting point for calculating energy expenditure of PoW based blockchains though.

[LoyceV]

I´m writing my bachelor´s thesis with the working Title:
 Could Germany reduce its CO2 Balance by replacing coins and banknotes with Bitcoin?

I like where you're going, but I think you're starting from the wrong point. I'll give you some things to think about, even though it's not really what you were asking.

1. I can't imagine Germany has only 600 million financial transactions per year. That's only 7 transactions per person, or just 1 transaction every 7 weeks.

2. I think you're confusing "Bitcoin" and "crytocurrencies in general". You can't create your own Bitcoin, but you can create your own cryptocurrency.

3. Hashpower has nothing to do with the number of transactions. The high hash rate gives security against an attack on your network, but if going to be national, it can just as well be centralized. If you want something that can handle many transactions without high power requirements: Stellar Lumens for example can handle 1000 transactions per second. Note that creditcards process much more than that during Christmas shopping peaks, but that's for a global audience.

4. Why limit yourself to Germany? German (euro) cash can already be used abroad, and the same will happen to any cryptocurrency.

5. If you're looking at CO₂ reductions, don't forget the 80 million clients that consumers will be using! I can give anybody cash money without direct energy consumption, but to make an electronic payment I need to turn on devices.

6. If you want to explore other areas in your thesis, security might be a nice one. How are you going to keep funds secure when millions of old people, born when computers were "programmed by switches and plugs", are responsible for their own digital cash?

Last point: please post your thesis results here when you're done.

[buwaytress]

But if you're already set at 600m txs per year... the current Bitcoin protocol couldn't process 600m a year.

At best, Bitcoin can process about 7 txs a second (though I think even at peak efficiency it only managed around 4). That's only about 220m txs a year. Your blockchain would need to make blocks about 3 times faster.

That's TPS number before SegWit soft-fork/improvement is activated. While i couldn't find accurate number, if majority bitcoiner use SegWit, the maximum TPS on Bitcoin network could reach between 12-20 TPS.
Even so, Bitcoin isn't solution for OP's thesis even if we considering side-chain, off-chain or other 2nd layer solution such as LN, expect OP considering Germany government/central bank consider using decentralized cryptocurrency.

Reference : https://bitcointalk.org/index.php?topic=3590240.msg36497029#msg36497029

Correct, except the majority of transactions or most common are no longer the single output single input tx we used to see a lot of (also used in yr referenced calculation). Tx sizes are actually increasing, even with the efficiency of Segwit as you mentioned.

The unseen good side of that is that people are learning to make the most of Bitcoin... Batch spends instead of individual ones. Consolidating inputs. So yes, overall, we see a lot of increased efficiency but actually, that 12-20 tps is not achievable. Even the old 7 tps is not achievable with the average sizes of today's txs. As I said above, the BEST we have done since November (after segwit) was 4 tps when virtually every block was filled to max capacity.

[kaffeeladen]

Aside from reasons why your idea simply impossible with Bitcoin or any decentralized Blockchain, Blockchain as database is less efficient than most centralized database management.

Also, i don't see any reason German Government would want use decentralized currency.
I think switching their currency to digital currency and become cashless society is more realistic and this also can reduce CO2 and climate change if their system is done correctly.

Good Point! That is absolute right - but i would like to exclude govermental control as it is one of the key ideas.
The thesis is inspired by the whole CO2 issue which currently discussed.

[kaffeeladen]

I´m writing my bachelor´s thesis with the working Title:
 Could Germany reduce its CO2 Balance by replacing coins and banknotes with Bitcoin?

I like where you're going, but I think you're starting from the wrong point. I'll give you some things to think about, even though it's not really what you were asking.

1. I can't imagine Germany has only 600 million financial transactions per year. That's only 7 transactions per person, or just 1 transaction every 7 weeks.
1,5 transaction/person / day as the survey from the central bank is published.

2. I think you're confusing "Bitcoin" and "crytocurrencies in general". You can't create your own Bitcoin, but you can create your own cryptocurrency.
Ok, this i have clarify in the thesis, but i will use bitcoin catch attention.

3. Hashpower has nothing to do with the number of transactions. The high hash rate gives security against an attack on your network, but if going to be national, it can just as well be centralized. If you want something that can handle many transactions without high power requirements: Stellar Lumens for example can handle 1000 transactions per second. Note that creditcards process much more than that during Christmas shopping peaks, but that's for a global audience.

4. Why limit yourself to Germany? German (euro) cash can already be used abroad, and the same will happen to any cryptocurrency.
- limeted, by time, data availibility, so I need modell my self-

5. If you're looking at CO₂ reductions, don't forget the 80 million clients that consumers will be using! I can give anybody cash money without direct energy consumption, but to make an electronic payment I need to turn on devices.
Yes but money, has to be minted, stored, delivered eg. ATM, Light, etc, fraud and much more point. even carrying money is weight causing co2 emissions.

6. If you want to explore other areas in your thesis, security might be a nice one. How are you going to keep funds secure when millions of old people, born when computers were "programmed by switches and plugs", are responsible for their own digital cash?
- as studies already show, Germans are not realy a  nation leadusers- they want to stick to cash


Last point: please post your thesis results here when you're done.

[kaffeeladen]

How would you design/ calculate the setup ( cpu power) for national cryptocurrency- when 125´000´000 transaktion per day have to be executed? ( 11 nodes/ miners and alternatively with 30000 nodes/ miners)?


11 for the amount of national banks´branches -
30000 for branches of the private bank sector

[zadelmohtarif]

You can't calculate the exact hashpower that would be needed.
Blockchain adjusts the required hashrate about once in every 2 weeks. That means that regardless of the used hashrate about 1 new block is found in 10 minutes (on average).

[HeRetiK]

How would you design/ calculate the setup ( cpu power) for national cryptocurrency- when 125´000´000 transaktion per day have to be executed? ( 11 nodes/ miners and alternatively with 30000 nodes/ miners)?


11 for the amount of national banks´branches -
30000 for branches of the private bank sector

As mentioned above, the required computation power for securing a blockchain has nothing to do with its transaction throughput. You should really do more research on what PoW actually does in the context of cryptocurrencies before jumping to the faulty conclusion that it directly correlates to transaction throughput and scalability.