4 ms·

How many joules per transaction is the proof of stake network? How many tps can it perform?

DennisP 5y ago
Based on observations, it looks like the proof of stake network will drop Ethereum's energy usage by about 99.95%. https://blog.ethereum.org/2021/05/18/country-power-no-more/ https://blog.ethereum.org/2021/05/18/country-power-no-more/ It can do this because it doesn't rely on any sort of "work." It's just another internet protocol. Proof of stake doesn't have anything to do with tps, so that won't change significantly. The solutions to that are rollups (second layers that inherit the full security of the main chain) and sharding (which multiplies the capacity of rollups). Some rollup systems are in production today, capable of a couple thousand tps. According to the most recent plan I saw, sharding would multiply that by twenty, initially.
magicjosh 5y ago
In terms of "tps", transactions per second, Proof of Work Ethereum can handle 30 transactions per second. Part of the development for "Proof of Stake" Ethereum includes scaling the number of transactions via sharding. Each shard will be able to handle 2,400 tps. As more shards are deployed up to 64, Ethereum will reach 160,000 tps. Please note, the upcoming improvement to Proof of Stake Ethereum does not yet use sharding, but it is under development. As comparison 1,700 tps for the Visa network is a good benchmark. I recall Visa's technical capability is 10x or 100x that number though. This comparison isn't intended to claim "ETH beats Visa for currency stuff", but simply to give a sense of the potential impact of Ethereum as it improves. Thousands of transactions per second implies a powerful network, whatever it's used for.
- rglullis 5y ago
  These claims for Ethereum doing 100k tps on the base layer are bogus. It might look nice, but completely out of reality from a fully distributed network. None of the core devs talk about these things anymore and everyone has accepted that the way to scale will involve layer-2 systems and some sacrifices in decentralization for performance. Anyone claiming to do 1k+ tps on a blockchain is either lying or just running things on a very limited or expensive network. Even then, take a look at Solana and see how many outages they are having lately.
  - magicjosh 5y ago
    Hmm source? Unless something has changed since January 3rd, 2022, sharding is intended as I stated. Here's an interview with Vitalik one of the core devs of ETH: https://youtu.be/b1m_PTVxD-s?t=1466 https://youtu.be/b1m_PTVxD-s?t=1466 Or perhaps you're saying Vitalik isn't relevant anymore? Perhaps he's only a figurehead now or something. Or perhaps I have something more to learn about how this all works. Again, sources would be great, I'm curious for what the future holds.
    - rglullis 5y ago
      I am not questioning the roadmap, or that they want to have sharding. I am questioning the numbers. Getting 2400tps per shard is already a pipe-dream, but tps does not scale linearly with the number of shards. The more shards you have, the more transactions will require cross-shard communication, bringing total tps down. Also, the video you linked has Vitalik talking about roll-ups as the main way to scale. With roll-ups, sure, we can get to 100k tps, but like I said in this case we are not talking about base-layer anymore.
      - magicjosh 5y ago
        Ah ok I'm getting sloppy with the terms roll ups and shards and I fundamentally understand neither. Happy to read any articles you wish to share on this. You mentioned that the core devs have changed their conclusions, for example, would be great to hear what they're thinking.
        rglullis 5y ago
        > I fundamentally understand neither. Then, please-pretty-please do not put yourself in a position of spreading misinformation. Even if unintentionally. It's hard enough to fight the stigma even when armed with good information, when "critics" see unfounded claims like yours it confirms the bias against crypto even more. > Happy to read any articles No articles. Stay away from opinion pieces. Try to get actual documentation about the project. Ethhub has some good material about it (https://docs.ethhub.io/ethereum-roadmap/ethereum-2.0/eth-2.0-phases/ https://docs.ethhub.io/ethereum-roadmap/ethereum-2.0/eth-2.0...). Look at the "Layer-2 scaling" section to learn about roll-ups and the many other solutions.
rglullis 5y ago
Ethereum plans to scale is not just on the switch to PoS. It will also include layer-2 systems. In any case, currently there are ~300000 validator nodes running the beacon chain, which does not require powerful hardware. For comparison, 13.4M units of the Playstation 5 were sold since 2020. 116.6M units of the PS4 were sold since 2013. I'd like to know the number of servers runnning Minecraft there are around, but you can unarguably say that people waste more energy playing videogames than with the ethereum network.
- inlined 5y ago
  PS5 uses a peak of 200W[0]. If they were used constantly, it would be 200Wh/h or 1.75MWh/yr. 13.4M PS5s would kick that up to 23TWhr/yr. The energy use of Eth is estimated at over 100TWHr/yr[1]. So no… even if every PS5, a chip that has climate scientists worried, that was ever created was left doing something horribly inefficient (say, computing PoW coin) 24x7 it would be using a quarter of this one coin’s energy consumption. [0]: https://www.gamesradar.com/why-climate-scientists-think-the-ps5-and-xbox-series-x-are-still-too-much-of-an-energy-drain/#:~:text=A%20new%20report%20from%20the,energy%20usage%20of%20their%20predecessors https://www.gamesradar.com/why-climate-scientists-think-the-.... [1]: https://www.statista.com/statistics/1265897/worldwide-ethereum-energy-consumption/ https://www.statista.com/statistics/1265897/worldwide-ethere...
  - rglullis 5y ago
    Forget about PoW. I was talking about the 300k PoS validators on the Beacon chain ("ETH2.0"). You can run a validator on less than 10W. Any cheap celeron can do it.
landemva 5y ago
I wonder if we should rate USPS by joules per envelope delivered.

Based on observations, it looks like the proof of stake network will drop Ethereum's energy usage by about 99.95%. https://blog.ethereum.org/2021/05/18/country-power-no-more/ https://blog.ethereum.org/2021/05/18/country-power-no-more/ It can do this because it doesn't rely on any sort of "work." It's just another internet protocol. Proof of stake doesn't have anything to do with tps, so that won't change significantly. The solutions to that are rollups (second layers that inherit the full security of the main chain) and sharding (which multiplies the capacity of rollups). Some rollup systems are in production today, capable of a couple thousand tps. According to the most recent plan I saw, sharding would multiply that by twenty, initially.

magicjosh 5y ago

In terms of "tps", transactions per second, Proof of Work Ethereum can handle 30 transactions per second. Part of the development for "Proof of Stake" Ethereum includes scaling the number of transactions via sharding. Each shard will be able to handle 2,400 tps. As more shards are deployed up to 64, Ethereum will reach 160,000 tps. Please note, the upcoming improvement to Proof of Stake Ethereum does not yet use sharding, but it is under development. As comparison 1,700 tps for the Visa network is a good benchmark. I recall Visa's technical capability is 10x or 100x that number though. This comparison isn't intended to claim "ETH beats Visa for currency stuff", but simply to give a sense of the potential impact of Ethereum as it improves. Thousands of transactions per second implies a powerful network, whatever it's used for.

rglullis 5y ago
These claims for Ethereum doing 100k tps on the base layer are bogus. It might look nice, but completely out of reality from a fully distributed network. None of the core devs talk about these things anymore and everyone has accepted that the way to scale will involve layer-2 systems and some sacrifices in decentralization for performance. Anyone claiming to do 1k+ tps on a blockchain is either lying or just running things on a very limited or expensive network. Even then, take a look at Solana and see how many outages they are having lately.
- magicjosh 5y ago
  Hmm source? Unless something has changed since January 3rd, 2022, sharding is intended as I stated. Here's an interview with Vitalik one of the core devs of ETH: https://youtu.be/b1m_PTVxD-s?t=1466 https://youtu.be/b1m_PTVxD-s?t=1466 Or perhaps you're saying Vitalik isn't relevant anymore? Perhaps he's only a figurehead now or something. Or perhaps I have something more to learn about how this all works. Again, sources would be great, I'm curious for what the future holds.
  - rglullis 5y ago
    I am not questioning the roadmap, or that they want to have sharding. I am questioning the numbers. Getting 2400tps per shard is already a pipe-dream, but tps does not scale linearly with the number of shards. The more shards you have, the more transactions will require cross-shard communication, bringing total tps down. Also, the video you linked has Vitalik talking about roll-ups as the main way to scale. With roll-ups, sure, we can get to 100k tps, but like I said in this case we are not talking about base-layer anymore.
    - magicjosh 5y ago
      Ah ok I'm getting sloppy with the terms roll ups and shards and I fundamentally understand neither. Happy to read any articles you wish to share on this. You mentioned that the core devs have changed their conclusions, for example, would be great to hear what they're thinking.
      - rglullis 5y ago
        > I fundamentally understand neither. Then, please-pretty-please do not put yourself in a position of spreading misinformation. Even if unintentionally. It's hard enough to fight the stigma even when armed with good information, when "critics" see unfounded claims like yours it confirms the bias against crypto even more. > Happy to read any articles No articles. Stay away from opinion pieces. Try to get actual documentation about the project. Ethhub has some good material about it (https://docs.ethhub.io/ethereum-roadmap/ethereum-2.0/eth-2.0-phases/ https://docs.ethhub.io/ethereum-roadmap/ethereum-2.0/eth-2.0...). Look at the "Layer-2 scaling" section to learn about roll-ups and the many other solutions.

rglullis 5y ago

Ethereum plans to scale is not just on the switch to PoS. It will also include layer-2 systems. In any case, currently there are ~300000 validator nodes running the beacon chain, which does not require powerful hardware. For comparison, 13.4M units of the Playstation 5 were sold since 2020. 116.6M units of the PS4 were sold since 2013. I'd like to know the number of servers runnning Minecraft there are around, but you can unarguably say that people waste more energy playing videogames than with the ethereum network.

inlined 5y ago
PS5 uses a peak of 200W[0]. If they were used constantly, it would be 200Wh/h or 1.75MWh/yr. 13.4M PS5s would kick that up to 23TWhr/yr. The energy use of Eth is estimated at over 100TWHr/yr[1]. So no… even if every PS5, a chip that has climate scientists worried, that was ever created was left doing something horribly inefficient (say, computing PoW coin) 24x7 it would be using a quarter of this one coin’s energy consumption. [0]: https://www.gamesradar.com/why-climate-scientists-think-the-ps5-and-xbox-series-x-are-still-too-much-of-an-energy-drain/#:~:text=A%20new%20report%20from%20the,energy%20usage%20of%20their%20predecessors https://www.gamesradar.com/why-climate-scientists-think-the-.... [1]: https://www.statista.com/statistics/1265897/worldwide-ethereum-energy-consumption/ https://www.statista.com/statistics/1265897/worldwide-ethere...
- rglullis 5y ago
  Forget about PoW. I was talking about the 300k PoS validators on the Beacon chain ("ETH2.0"). You can run a validator on less than 10W. Any cheap celeron can do it.

landemva 5y ago

I wonder if we should rate USPS by joules per envelope delivered.