How do miners profit post-PoS, if at all, how will new miners get their start, and who will execute the EVM byte code (and why)?

  • I've been following the release of Casper/PoS closely, and I'm curious about a few details.

    1. How will miners make money (even if GPUs are less valuable)? Are spare compute resources (say RAM, CPU) still somehow "usable" to create value by securing the block chain in a PoS world?
    2. How will miners new to Ethereum gain a foothold if they start with no stake and have no ability to mine their first ETH coin?
    3. If spare computing power isn't the resource that drives mining, who will be incentivized to participate? Will the mining/consensus system be entirely driven by users of the network, who hold ETH for other reasons entirely (e.g. users of ETH-powered contracts)? And if so, how can they simultaneously use their ETH to power contracts, collateralize contracts (potentially), and simultaneously serve as a "deposit" for Casper?
    4. Who will execute the contracts/EVM bytecode?
    5. Does PoS make computing power no longer "redeemable" for financial value using the Ethereum block chain?

    Thanks very much, and looking forward to the next gen of Ethereum!

    Welcome to Ethereum Stack Exchange! It is preferred if you can post separate questions instead of combining your questions into one. That way, it helps the people answering your question and also others hunting for at least one of your questions. Thanks!

    The 5 parts of his questions are so closely related and are different aspects of the same thing that I doesn't make me feel bad to see it in a single question. So I won't recommand splitting.

    I read but not clear. After POS, how to create new blocks and who will do it when all miners to stakers?

    Will someone staking say 100 - 200 ETH have any incentive under this system? Or will it only be worth while to those holding thousands or tens of thousands? Once all ETH has been released/mined, where will the ETH come from to reward those staking under PoS? Thanks in advance for clarifying

    1. Miners will receive transaction fees, along with (potentially) a small block subsidy. GPUs will no longer be useful, and CPU/RAM will be useful in that stakers will need to have close to perfect uptime and fast servers in order to make money.

    2. They will have to buy ETH, and stake it.

    3. The stakers will have to put down a deposit(stake), that cannot be used for other purposes while it is staked. Stakers will likely be professionals, much like how mining pools are now.

    4. Similar to today, the miners/stakers will execute the contracts and build the blocks, and each node will verify the computation in order to be sure the block is valid.

    5. Essentially, yes, although fast servers are needed for processing transactions at the higher rate that Casper allows.

    4. except there is no need for nonce anymore is it. and THAT, saves the planet.

  • EDIT: This answer is a bit old. Now we know much more about PoS and validation details, but don't hesitate to ask new questions on a new thread.

    Also, meanwhile Golem Network appeared and is a pretty way to recycle GPUs power (see the end of my answer where I express fear about GPU recycling). GPUs would then be used for what they were built for: rendering and massive calculation. Many other projects may appear and I'm happy that my prediction about people being creative were true ;)


    TL,DR: I may look pessimistic about POS but there is a huge chance that people creativity makes something good of it. However there will be a huge amount of GPU to recycle, so we have to find a solution now.


    Now the long talk. As I can see it, POS transition was primarily started to lower the ethereum network power consumption and be more ecological. But the drawback is that tons of GPU will be left useless (even if there is still some POW coins out there, but come on, mining LiteCoin to convert it to ether?) and trashed if not sold. I really mean tons. Really ecological? One may do the math but I don't have the figures to put in the formula. Also POW mining looks to be a more accessible, decentralized and incentive way to be part of a network you care about. Will you be really a part of something ruled by a bunch of big firms that hold ( or hodl) a lot of ether? I think this is the opposite of what ethereum was build for first. But, answering your questions:

    1- as I understood it, what we call miners today will be validators. They will exist in limited number as POS is described for now but this may change. So not every people having a node will be rewarded. Only a few big corporations with a lot of stake to participate in POS will be able to be rewarded. The more you have, the more you gain... All our rigs will be useless even for being recycled into POS servers as the need is really different and don't expect to build a high availability powerful server on an Asus btc motherboard. So except if you want to mine some other coins or sell your rig, this is good to trash. (Note that selling rigs and GPU will be nearly impossible when POS will be there due to the gigantic gpu offering opposed to the tiny demand as none will want POW hardware anymore)

    2- new miners will have some way to get ether. First one is to buy some at an exchange using fiat currency. This is how most of non miners do today and this is how I stared before being a mining addict :D. They will also be able to change other coins like BTC or LTC they already have into Ether using an exchange or the now famous and beautiful ShapeShift (I love you guys even if all this hack thing is not completely clear to me... But this is another thread subject). Note that if a miner don't want to use fiat currency at all, he can mine LTC for instance and then change it to ether. But this is not really efficient, I admit. There's some other way though. Some can sell goods or services for ether. I already saw a young designer offering logo creation on Reddit for ether. Gamblers will also be able to win some ether by playing online games like poker. Even if they'll have to put some at first, they will be able to win or loose a lot. Finally (but not so finally as I may have not thought about anything), you could be rewarded in ether by watching ads or solving captchas for hackers, this is called faucets and I hope it will be eradicated.

    3- mining, or may I write validating, will be the only way to create ether from void. So in order to keep the network ether creation rate, POS will have to be attractive. This will be done by giving a few people the ability to invest ether at a really good reward rate (most classical investment in old world won't be able to compete with POS rates). The more they stake, the more they get. As staking ether will lock it for some times, it will probably be people who owns a lot or people who don't want to use it (Holders). But people who have just some ether and use it won't be part of it even if the network works because of them and they create the ether value by using it and using Dapps and contracts. But they are de facto excluded form the reward. So yes, you will be able to stake and hold and use it if you have enough ether.

    4- as I understood it, validators with their huge servers will process all. Full nodes that some people will run will only validate blocks integrity but won't run any EVM code or create new blocks. This is what happens today with full nodes in POW. Only mining nodes run the network. I don't know how light nodes will play in this game. They may only have a read and write access but no integrity validating power at all. Read the light protocol paper for more details.

    5- computing power will still be redeemable but not as it is today. GPU power will be almost zero value. But computing power on super servers or on big arrays, grids clusters will be very valuable.

    Now the optimistic point:

    By that I mean that what would be a more available computer than some standard computers joining together to make a big one always on? This may be the future POS pools. People will join their CPU power using some networked clustering tool and each one will stake some ether in a common wallet contract or some forms of DAO. They will then apply to become a validator as one could be and they will be represented by the contract. The contract will receive rewards and distribute it proportionally to the computing power delivered and by the staked amount of each people as it's done in pools with hashrate.

    This is the best scenario I see to avoid power centralisation of the ethereum network.

    And the pessimistic message for the end.

    But all our GPUs will go to trash anyway.

    "We can't switch over to cars - what will we do with all these buggy whips?"

    Cars and buggy are not planned to be planet friendly as I now. Here, the question of the waste generated by POS is a real matter in order to build a better decentralized world. This should not be done with a huge ecological depth.

    And by the way whips are only wood and leather... GPU are highly toxic materials with a very long life and recycling them costs a lot of energy and generates a lot of polution in mainly poor countries as India or China. So if you don't care, I do.

    They're already made and purchased. Changing to a system that doesn't require them doesn't waste any resources that weren't already wasted - and it prevents further waste. Unless you thought that nobody else will ever add any more processing power than there is now?

    Yes but even it they are already made, if none needs it, they will be trashed. When they are running in your rigs, they are not creating more pollution in the river nearby the recycling zone in a poor country. Building them already generated a lot of wastes, so we should not let them generate more...

    ... (sequel fo my previous comment) and when you say that they are already built, you just ignore the fact that a huge part of the cards were built because of the mining demand. Otherwise do you think AMD would have built so much cards only for gamers ? Do you think they are so bad in calculating gamers GPU market needs that they built 50 times more GPU than necessary ? Come on...

    And I don't say that we should not move to POS, I just say that we have to do it the right way by taking the miners and existing hardware in account not to generate too much pollution and deceived people.

    They'll still be built, but they won't be consuming electricity. Why do you assume that the only cost is manufacturing? And how could inventing some make-work for the cards to do, again consuming electricity, possibly be better than not using them at all? They won't get any less toxic if you throw them away in 3 years when they're obsolete than if you sell them on to gamers now.

    Do you really think miners who owns dozens of cars will only keep them in a box not to pollute ? Do you think the gaming used GPU market is ready to absorb such a quantity of cards ? Do you think gamers will buy cards that where OC and reached 70°C 24/7 ? I don't. It's however interesting that you talk about a make work project. Do you have any clues on what it could be ? If we could find a way to make something big and useful with all these GPUs this would be nice. This would be a nice way to buy back GPUs (with Ether) and make miners happy.

    Perhaps you can clarify why having them hang around consuming electricity for a few more years before being disposed of anyway is an improvement, then?

    First I'm not saying we have to keep POW and let all those cards running and consume electricity. POS is nice and clearly needed. I only say that we have to care about what all these cards will become. That's all.

    Then I don't understand your point. The cards already purchased are a sunk cost. Inventing some way for them to do something won't reduce their environmental impact - it'll increase it.

    Yes, exactly. We agree on this. Don't try to make me look like I don't agree with you, I do ! But throwing them to trash would be an even more disaster and selling them is not realistic. Keeping them in a box at home is not an option as people need spaces (perhaps for a regular heater, now that they don't have their rig to heat the house). What should we do ? You Google guys should have nice ideas about this environmental problems don't you ?

    I am new to all this but .. What about the possibility of POS pools? It is a risk locking up x number of ETH with a centralized pool. Perhaps POS works with smart contracts, where the ETH get locked in a contract for specific amounts of time, the contracts are what is submitted to a pool. A pool is made up of may contracts all with a variable time rates. Each contract earning percentage of reward.

    @farinspace Yes it's exactly what I wrote in my answer if you read after "the optimistic point"

    NMassart, indeed, sorry about that, lol ... at least, it feels good that my thinking is along the lines of the experts!

    NMassart, your thoughts on creating a decentralized AWS like GPU focused arrays/clusters with pricing rewards for those running nodes? ... so as not to have GPUs going into the trash

    @farinspace Sorry if I misunderstood your comment, I'm not an English native. But in my answer I write about CPU array, not GPU. We all agree that GPU power consumption and heat are both things we have to get rid of. And by the way I'm far from being an expert... really.

  • PoS will be the saving force behind decentralization. Currently the wastage is too large to be sustainable. It is currently being driven by speculation.

    The idea of mining was to secure the network and it has become a delicate business involving much risk. As an investor into mining technology you should be aware of the implications of PoS.

    Bitcoin and all decentralized tech is not about mining it is about the application and usage of that technology and any saving in efficiency can only help it along. "Mining" will now have to shift toward providing more resources to the network and less towards security as the security issues (in theory) are taken care of with PoS, mining will make a lot more sense.

License under CC-BY-SA with attribution


Content dated before 7/24/2021 11:53 AM