A useful PoW without replacing Nakamoto Consensus

[kernel1983]

Hi, this is KJ. I am a Ph.D. candidate from the University of Liverpool.

I am proposing a useful PoW without replacing Nakamoto Consensus. I named it Economic Proof of Work (EcoPoW).
I have talked about this algorithm to several experts in Blockchain,
including Dr. Jiaping Wang, Dr. Qi Chen from MSRA, Dr. XiaoKang Mo, Prof. Xi Chen from NYU, and received positive feedback.
It is relatively simple to understand if you are familiar with PoW and Proof of Replication (PoRep).

EcoPoW uses PoW as an encoding algorithm during the PoW outputs consensus.
As the consensus uses the left bits of the hash output binary (the more 0s, the more difficulty), the encoding algorithm uses the right N bits.
When N=8, we encode 1 byte at a time (around 256 sha2 hashes to encode 8 bits).
Since N is adjustable, we can change the difficulty of encoding, such as 2 bytes a time (around 65536 hashes to encode 16 bits ).
The encoding is expensive as designed, and decoding is cheap as it only takes one hash operation.

The reason we need to use PoW as an encoding algorithm is that the computation can be used for Proof of Replication.
It is the key algorithm for blockchain storage, such as Filecoin.
Filecoin uses PoRep for user data security against Outsourcing Attacks.
The core idea of PoRep is to encode the user file's original content into a unique replication with their unique name as a key (a public key can be used as the key).

When verifying, the prover needs to compute the proof from the replication instead of the original content.
Even the prover can perform an Outsourcing Attack by fetching the replication from others and decoding in seconds,
and he still needs a lot of computation and time to encode the decoded content with his name into his unique replication.
This prevents the Outsourcing Attacks as it is cheaper to keep the replication honestly on his storage device.

After diving into the research, once the PoW is useful, there are some important impacts:
1. The miners are making money by selling their computation and storage resources.
The coins are not the only revenue source for the miners.
2. Once the computation is consumed locally, miners will no longer sell the computation to the mining pool to exchange the coins.
It will bring true decentralization to the blockchain world.
3. The PoW computation is task-based.
If there is no demand, the computer will go to sleep mode to save energy.

We propose to build a blockchain with storage and EcoPoW.

[1714290350]

1714290350

[1714290350]

1714290350

[1714290350]

1714290350

[NotATether]

3. The PoW computation is task-based.
If there is no demand, the computer will go to sleep mode to save energy.

I think the last part here is key to building a proof-of-work based consensus without all the energy consumption.

At some point in the blockchain economy, when the (logarithmic) price curve crosses the corresponding curve for difficulty - possibly with some adjustments because price and difficulty are not 1:1 - it will make no sense to throw additional hashpower to the network.

So when such a situation happens, which I'm guessing will be after a few decades, ASIC vendors can include a "sleep mode" with Wake-on-LAN feature so that a pool can wake up the miner when the difficulty reaches a certain threshold.

My suggestion won't lower existing energy usage, but puts a "soft cap" on the energy use without the need for a soft fork.


Is there a link to the research paper of EcoPoW?

[NeuroticFish]

I am unsure though what would stop somebody change his node to still accept (and then rely) new blocks even in the moments he should be sleeping.
I don't know if this can be done without a centralized entity that confirms the blocks were created only when they should (or sends out the tasks and "pays" for the blocks).
What am I missing?

[kernel1983]

OK, I cannot estimate how much energy would be saved, or more would be wasted, unless we have this chain running.

However, even more energy were burned, now it is not for the chain security only, the computation is used for a honest proof of the user data storage.

I think it is risky/difficulty to change Bitcoin into a storage blockchain now. But Im planning to make a storage chain with this design.

Beside, the tasks are the storage requests. In this era, we generates data any moment. So ideal there are enough tasks making the blockchain security.

And yes, more feedbacks, please.

[kernel1983]

No, I'm not selling this to Bitcoin for any hard fork.

The idea is about how we can make PoW useful.
It is too risky and too big change for Bitcoin Network, at least for now.

However, it is still interesting if EcoPoW can be implemented. I'm working on it.
Because I see the opportunity that we can have a PoW blockchain without a mining pool.
That makes me excited.

As ETH left PoW, BTC is the only landmark for decentralization now.
However, the mining pool is still there. Maybe it can be removed?

If this works, we would have a more decentralized blockchain.

[BlackHatCoiner]

For those who don't know what's Proof-of-Replication: https://filecoin.io/proof-of-replication.pdf

It is relatively simple to understand if you are familiar with PoW and Proof of Replication (PoRep).

I took a quick glimpse to Proof-of-Replication, and the paper essentially describes a way that nodes can prove the replication of their dedicated physical storage? Sort like Proof-of-Space Chia Network uses?

The idea is about how we can make PoW useful.

This particular sentence says a lot. You have it wrong. Proof-of-Work is already useful. It protects the network from unreliable factors. It's the, yet, unique solution to Byzantine's fault.

[NotATether]

OK, I cannot estimate how much energy would be saved, or more would be wasted, unless we have this chain running.

However, even more energy were burned, now it is not for the chain security only, the computation is used for a honest proof of the user data storage.

I think it is risky/difficulty to change Bitcoin into a storage blockchain now. But Im planning to make a storage chain with this design.

Beside, the tasks are the storage requests. In this era, we generates data any moment. So ideal there are enough tasks making the blockchain security.

And yes, more feedbacks, please.

I also believe its better to make a "pilot" storage blockchain because the chances of changing the underlying consensus type of Bitcoin is practically zero, for the distant future.

What I suggest is the following:

Try to emulate transaction volume between January 2021 and January 2022 on the pilot blockchain, so you can make an accurate estimation of the results or whatever you want to measure: efficiency, energy, or whatever.

However, it is still interesting if EcoPoW can be implemented. I'm working on it.
Because I see the opportunity that we can have a PoW blockchain without a mining pool.
That makes me excited.

I'm pretty sure that PoW is already useful in that it accomplishes what Nakamoto2008 described for it to do, but I am curious to know how a miner-less blockchain would look like in terms of participants. It is not desirable for the main stakeholders to be exchanges and liquidity pools (Proof of Stake / ETH) or banks (XRP).

[BlackHatCoiner]

Also, take into consideration that you can build stuff with sidechains. No consensus is required, and it is based on the Bitcoin blockchain.

I'm pretty sure that PoW is already useful in that it accomplishes what Nakamoto2008 described for it to do, but I am curious to know how a miner-less blockchain would look like in terms of participants.

Me too. What I'm confident about is that a mechanism with a real-world cost parameter missing comes with great disadvantages.
 

(Proof of Stake / ETH) or banks (XRP).

It's pretty much the same thing. Bankers already work as stakeholders.

[n0nce]

OK, I cannot estimate how much energy would be saved, or more would be wasted, unless we have this chain running.

However, even more energy were burned, now it is not for the chain security only, the computation is used for a honest proof of the user data storage.

I would like to comment on 2 subjects that you also mentioned here.

First of all, conceptually, when mining Bitcoin, the energy is not wasted. Nobody would say that a car factory wastes energy to run their machines either, right? You are minting new coins and expending energy for that; I don't understand why so many people believe this is wrong.

Secondly, also conceptually, adding a secondary use to your mining, such as using the heat or using an algorithm that performs 'useful' computations - since all of this has monetary value, you just extract more money out of the same energy, i.e. just increasing the efficiency of your gear. This means everyone will do it, the power consumption will not drop and the difficulty just goes up a little.

[kernel1983]

I totally agree that PoW is useful.
However, the others who are worried about the energy cost of PoW, most of them choose PoS now.
I wouldn't say I like PoS, as the soul of blockchain, is about the miners contributing resources but not capitalizing.

So, another direction we can try is to make PoW "more useful".
Assuming there are three roles:
Blockchain system, miners, and users in the system,
I can say the current PoW mining is useful to the blockchain system security.
The user cannot feel insecure unless the amount of computation is low enough (most of the miners quit).

EcoPoW is adding another usage to the existing PoW.
It did not change PoW for consensus at all.
And the PoRep/encoding with EcoPoW does not cost extra computing.
It is useful to blockchain users as proof that the data is kept by a resource (storage) provider honestly.
The storage providers (miners) can charge the users with the proof. It is the new revenue source for the miners.


The Proof of Replication and Proof of Space is different.

Proof of Replication (PoRep) does not generate consensus.
It only shows the data were kept by another user, honestly (on a dedicated storage, to prevent an OutSourcing Attack).
It saves useful user data on disk.
Filecoin uses PoRep for user data security and PoS for consensus.

Proof of Space (PoSpace/PoC) is a replacement for PoW.
It outputs consensus but no useful user data on disk.
It uses computation to pre-generate a lot of trash data on disk.
After pre-generation, there are less energy cost for the consensus generation.
Chia project uses PoSpace so it is not a blockchain storage project.

[NotATether]

I wouldn't say I like PoS, as the soul of blockchain, is about the miners contributing resources but not capitalizing.

I agree with this.

So, another direction we can try is to make PoW "more useful".

I'm not going to get into any more bikeshedding about usefulness so let's just leave it at that.

Assuming there are three roles:
Blockchain system, miners, and users in the system,
I can say the current PoW mining is useful to the blockchain system security.
The user cannot feel insecure unless the amount of computation is low enough (most of the miners quit).

EcoPoW is adding another usage to the existing PoW.
It did not change PoW for consensus at all.
And the PoRep/encoding with EcoPoW does not cost extra computing.
It is useful to blockchain users as proof that the data is kept by a resource (storage) provider honestly.
The storage providers (miners) can charge the users with the proof. It is the new revenue source for the miners.

I'm assuming this is going to work as long as a miner has other resources to sell such as storage, and by that metric, things like network traffic and RAM?

If so, then you should know that most ASICs only have a minimal amount of storage, so the EcoPoW cannot be directly used by them without dramatic redesigns of the ASIC. You can't just slap in a hard disk drive without shooting up the average cost of the miner by some margin.

[BlackHatCoiner]

However, the others who are worried about the energy cost of PoW, most of them choose PoS now.

The "others" who are "worried" about the energy cost of Proof-of-Work aren't sufficiently educated enough to understand Proof-of-Work, or they're just enemies of the ecosystem. We all know how Proof-of-Stake works, and if there were no tradeoffs compared to Proof-of-Work, we'd have already switched to it. But, it comes with some serious disadvantages, those are centralization and security sacrifice.

Anyway, if your idea requires no consensus, as I've already said, it can be built as a sidechain.

[kernel1983]

If so, then you should know that most ASICs only have a minimal amount of storage, so the EcoPoW cannot be directly used by them without dramatic redesigns of the ASIC. You can't just slap in a hard disk drive without shooting up the average cost of the miner by some margin.

Yes, let's go back to the PC and maybe GPU. The original idea of Bitcoin and PoW are using the spare computation of PC, right?

I'm assuming this is going to work as long as a miner has other resources to sell such as storage, and by that metric, things like network traffic and RAM?

PC have free bandwidth and spare compuation. The only cost is a 8T-16T hard disk. No SSD reqiuired.
And remember the user storage requirement is not one-time paid. There is sustainable incoming for miners.

The "others" who are "worried" about the energy cost of Proof-of-Work aren't sufficiently educated enough to understand Proof-of-Work, or they're just enemies of the ecosystem.

PoW is almost perfect except the energy cost and the mining pool. If the "enemies" are using the PoW computation, they would have nothing to complain.
If PoW still has space to improve, let's upgrade it! (Upgrade PoW first, not Bitcoin)
I do this because I love PoW.

We all know how Proof-of-Stake works, and if there were no tradeoffs compared to Proof-of-Work, we'd have already switched to it. But, it comes with some serious disadvantages, those are centralization and security sacrifice.

For decentralization, you definely want a PoW with no mining pool.

Anyway, if your idea requires no consensus, as I've already said, it can be built as a sidechain.

The core idea is let PoW algorithm output both consensus and Proof of Replication/encoding within same computation.
If we just use PoW for encoding, it make no sense to build anything.

[BlackHatCoiner]

However it didn't gain traction (probably because you need to run full node)

Or just because there isn't much concern. The game theory discourages all kind of attacks. Not just of an outside attacker.

Yes, let's go back to the PC and maybe GPU. The original idea of Bitcoin and PoW are using the spare computation of PC, right?

That wasn't the original idea, but it does form a more decentralized network, yeah.

If the "enemies" are using the PoW computation, they would have nothing to complain.

I'm not sure I make sense from this. What do you mean "to complain"?

For decentralization, you definely want a PoW with no mining pool.

Mining pools are necessary for low-capital miners to have a steadier income. What's needed is to discourage centralized pools' malicious reorgs further.

[HeRetiK]

PoW is almost perfect except the energy cost and the mining pool.

The energy cost is integral to PoW-based consensus security. Diverting part of the energy cost to "useful" computation, takes away from the security of the network as means of payment. But maybe this property is not as fundamental as it seems, so it will be interesting to see a fleshed out proposal on how to approach this matter from a different angle.

I'm not a fan of the necessity of mining pools for stable mining income either; however given the free choice and ease of switching mining pools it's less of an issue than it might appear to be at the first glance.

[Zilon]

After diving into the research, once the PoW is useful, there are some important impacts:
1. The miners are making money by selling their computation and storage resources.
The coins are not the only revenue source for the miners.

Who pays for the computation and storage resources been sold by the miners?

2. Once the computation is consumed locally, miners will no longer sell the computation to the mining pool to exchange the coins.
It will bring true decentralization to the blockchain world.

Is the coin been exchanged or is it serving as a new UXTO using the usual p2p system in a new address, then  secondly if miners no longer sell the computation won't it take even longer for confirmation so as to verify the legitimacy of the coin.

3. The PoW computation is task-based.
If there is no demand, the computer will go to sleep mode to save energy

We propose to build a blockchain with storage and EcoPoW.

Sorry to ask is the computer sleeping as a central storage network, will there be no contribution from other miners to keep the the network running

[casinotester0001]

3. The PoW computation is task-based.
If there is no demand, the computer will go to sleep mode to save energy

We propose to build a blockchain with storage and EcoPoW.

Sorry to ask is the computer sleeping as a central storage network, will there be no contribution from other miners to keep the the network running

Some smart miners will create 'fake' demand 

[n0nce]

EcoPoW is adding another usage to the existing PoW.
It did not change PoW for consensus at all.
And the PoRep/encoding with EcoPoW does not cost extra computing.
It is useful to blockchain users as proof that the data is kept by a resource (storage) provider honestly.
The storage providers (miners) can charge the users with the proof. It is the new revenue source for the miners.

You are trying to 'get more out' while 'putting the same in', though. Thermodynamics do not change.

I'd also argue that adding any such large change to Bitcoin has the potential to create so much more harm than good, e.g. through bugs and security problems, that it is completely unviable to do. And I also heavily doubt that the people who believe coins should be created out of thin air (unlike literally any other good or product), will be happy / silenced with such a compromise solution, anyway.

[kernel1983]

If the "enemies" are using the PoW computation, they would have nothing to complain.

I'm not sure I make sense from this. What do you mean "to complain"?

If the "enemies" complain the energy cost due to PoW, they will have no reason to complain EcoPoW as the computation is comsumed for both the chain security and the user data safe.
Otherwise, we must shutdown all the data centers for energy-saving.

For decentralization, you definely want a PoW with no mining pool.

Mining pools are necessary for low-capital miners to have a steadier income. What's needed is to discourage centralized pools' malicious reorgs further.

True, but there are new revenue source for the miners by selling their storage and computation.

Who pays for the computation and storage resources been sold by the miners?

The users who want to storage their data in a blockchain storage network.

Is the coin been exchanged or is it serving as a new UXTO using the usual p2p system in a new address, then  secondly if miners no longer sell the computation won't it take even longer for confirmation so as to verify the legitimacy of the coin.

The "exchange" I used here, is to give the computation to the mining pool and get coin reward in return.
If miners are not connecting to mining pool, all they are solo mining. It still protect the PoW network and new blocks are generated.

Sorry to ask is the computer sleeping as a central storage network, will there be no contribution from other miners to keep the the network running

We assume lots of encoding tasks are given to different PCs all over the world.
While those PCs finished the tasks are into the saving mode, the others are still working.
In this era we generate data every moment. The more data into the blockchain storage, the safer the blockchain is.

Some smart miners will create 'fake' demand 

If the users pay, the tasks are not fake.
Especially when the coin's value or price is low, then the user payment for the storage become the main revenue.
Unlike the coin reward in the solo mining, you dont need to be very lucky to get incoming.

Current PoW mining (where miner connect to pool) only require small bandwidth. If they're expected to store data which could be retrieved anytime, some miner will need to upgrade their internet plan either because low FUP or they subscribe to cheapest plan.

Compare to 2008, we have better bandwidth at home.
For the user storage request, it requires to be proved (PoRep) within 24 hours or 1 week, depending on the system design.
There will be enough time for a prover to download the data.

I'd also argue that adding any such large change to Bitcoin has the potential to create so much more harm than good, e.g. through bugs and security problems, that it is completely unviable to do. And I also heavily doubt that the people who believe coins should be created out of thin air (unlike literally any other good or product), will be happy / silenced with such a compromise solution, anyway.

Again, I said I'm not selling this to Bitcoin. Yes, the change is huge and it is not proven.
We can see this idea as an upgrade of PoW. I'd rather build a new blockchain storage project to see how things going on.
This give me the flexible to change anything, which is not limited in a framework.

[BlackHatCoiner]

If the "enemies" complain the energy cost due to PoW

I'm just dropping this here, because it's relevant: Complains don't matter. Actions do. We're perpetually debunking this environmental propaganda years now. They don't have counter-arguments to most of our arguments. Switching to another mechanism (theoretically, I know) because of these complains is not going to happen, because we're already dealing these with strong arguments.

True, but there are new revenue source for the miners by selling their storage and computation.

What does this have to do with miners' steadier income?