Skip to content

Latest commit

 

History

History
97 lines (71 loc) · 14.7 KB

20220215_013.md

File metadata and controls

97 lines (71 loc) · 14.7 KB
Raw
name date time location length
ETC Community Call 013
2022-02-15
1500 UTC
Discord
30-60+ mins

Description

A casual voice chat to discuss ideas for ETC. All are welcome.

The ETC Discord can be joined at https://ethereumclassic.org/discord

Please join us in the #community-calls channel to ask questions or bring up topics.

Agenda

Spillover Topics

Status

Meeting Minutes

The weekly Ethereum Classic Community Call 013 was held on 2022.02.15 at 1500 UTC on Ethereum Classic Discord.

Overview

During this meeting, the chair announced the successful implementation of the Mystique Hard Fork and informed the participants that a new version of the Ethereum Classic website will be released. Henry presented his proposals for a dual algorithm approach. Several items such as ETH & ETC mining,ASIC& GPU rigs, Hybrid PoW, transition plan, rebranding of Ethereum Classic in future, etc. were discussed at length. The participants shared their ideas and views, the gist of which is as under;

Points

  • Two announcements

    • The chair announced that the latest hard-fork ETC Mystique has been completed successfully. Thanks to everyone's hard work, it's been a smooth transition. He expressed congratulations to everyone involved and to the Ethereum Classic network.
    • The next announcement is that the Ethereum Classic website is very close to being launched. There's been work in the last month on new content concerning the significance and advantages of the Ethereum Classic. Different from the old website, the new version will focus more on the introduction and rationale for why people should be interested in Ethereum Classic. To make a better website, there will be a special call next week to invite suggestions.

  • The state of ETC mining

    • ETH VS. ETC mining: Today, the supply chain is constrained. Henry made a comparison between ETH and ETC mining. The Ethereum hash rate has witnessed a massive increase from 400TH in July to 1000TH. However, The ETC hash rate has seen a very consistent increase from 8 TH to 25 TH. In addition, ETH has a higher hash rate mainly because it's more profitable. On a high-level comparison, ETH has a 40% more hash rate with 57% higher revenues. When June comes along, the earnings on ETC are going to drop further by another 20%. One thing needed to bear in mind is that it's the prediction for revenues not profits which vary a lot depending on many factors such as people's hardware and power rates.
    • Reasons for low ETC hash rate: The first reason is that ETC has lesser profits for miners compared with ETH as aforementioned. The second reason lies in the mostly older equipment on ETC. Henry compared relative Information and earnings from the various GPUs. He illustrated that the people mining ETC are mostly with the older GPUs, 4GB models, and are 4-6 years old. Typical GPU rigs have 6-10 GPU cards depending on the power supplies in the rigs. Newer GPUs earn up to 7.5x more with up to 5x greater HR.
    • ASIC VS GPU rig (6-8 GPU): The ASIC rigs are getting 3-10 times more performance with higher hash rate earnings. They are way more power-efficient, accounting for 5% less of expenses while GPU rigs make up 33% of expenses. The ASIC rigs are also more reliable and easier to operate, lowering the cost of hosting. Besides, the ASIC rigs have more space density utilization.

  • Post Merge Mining Economics

    • Overpopulation of GPU & ETH ASICs: As WCCTech predicts, there will be a huge overpopulation of hardware. Henry summarized GPU PoW rewards top 7 non-ETH GPU coins based on F2Pool. Henry did not list all the GPU coins because calculating all of them means a lot of work and the bottom of the top seven coins reached only $6.3 which means it doesn't contribute to the total. He believed that GPUs are going to be unprofitable and unprofitable mining will lead to the massive shutdown of hash rate, or people will liquidate hardware or look for alternatives to recover the cost.
    • Massive population of ETH hardware: Henry listed four types of GUPs. They are RX570 GPU, RX570 (GPU) Rig, VU37P FPGA, A10Pro (ASIC). Among them, FPGA and ASIC exceed the size of the hardware type deployed based on his calculations.
    • Cost of operating ETC mining hardware: According to statistics from CryptoCompare, the GPU rigs cost $630 a year while the ASIC rigs cost $710 a year. The ASIC rig is making almost four times what the GPU rig is making. When it comes to mining, a lot of miners committed to fixed contracts are paying for their power in advance. Typically, the contract period might be two years. A lot of them are financed at very high-interest rates. Meanwhile, selling used equipment can result in large losses because the ASIC rigs have no value for they can only mine ETC hash or ET hash. Therefore, the alternative is to look for other ways to recoup the money.
    • Rig support VS ETC price: Based on the data from F2pool, Henry stated that GPU mining is going to be unprofitable due to overpopulation. There is only one-third of ETH GPUs supported with 8x ETC. ASIC mining tends to be supportable if ETC doubles in price, but profits will be slim. It's conceivable that FPGAs and high-end GPUs may coexist with ASICs but profits will be minimal.

  • Hybrid PoW Discussion

    • Hybrid PoW for ETC (Conceptual Diagram): Henry split the PoW 50 / 50 between ETCHash and Keccak-256. This diagram makes the network impossible to attack because there's an equal barrier on both sides. This conceptual framework needs to be adapted for ETC because the original difficulty code on ETC should also be considered and branched to the PoW framework. It may also entail the contribution of the community to help finish that.
    • Case study of the new KDA blockchain: It shows how ASIC ramps on blockchains. Henry did not believe that all ASICs are evil by giving the example of Kadena(KDA). KDA ranked #9 with $230k PoW rewards daily, making itself a very profitable coin to mine. Another benefit is that KDA ramps very nicely. The hash rate followed the price of the coin. Its hash rate has experienced steady growth from 1.8 PH to 84 PH.
    • Impact on ETC network: Henry believed it's critical to get ahead of post merge with some security mechanism. The 50/50 split between ETCHash & Keccak-256 is a critical defense against the overpopulation of ETHash hardware. ETC will be the dominant chain on Keccak 256. Based on Grin and Kadena's experiences, FPGAs & newer GPUs may be more profitable mining Keccak -256 than ETCHash in the first 6-12 months. Establishing the ETC brand might also be an alternative and help ETC take over part of the ETH smart contract. The thing that matters is to have a plan ready before the merge rather than reacting to the merge.

  • Q&A

    • The viability of the double algorithm: There are equivalent to 30 million GPUs mining ETC. It's impossible to make money if one moves those 30 million from ETH to ETC as clearly evidenced on the chart (p.13). They will get dominated by ASICs. The dual approach will protect the interests of the community because it sets a place between the two algorithms and draws the balance against the overall overpopulation of hardware.
    • Less-centralized GPU mining: Centralization is a matter of opinion. Based on the statistics and availability, ASICs are highly centralized. If we don't do something, we will be subject to the hordes of GPUs. There will be huge overpopulation, and no one makes any money. Security and stability are at risk. For this reason, the community has to make a move. The hybrid wall builds in natural protection and builds a balance. The SHA3/ASIC can ramp again in a controlled manner. The example of Kadena shows that the new hardware will build loyalty among investors.
    • The development of ETC/SHA3/hybrid system: The approaches are to have both. 50/50 is a good starting point because it builds in the balance naturally. The community can decide over time whether to migrate more to SHA3. We prefer moving more to SHA3 which gives better control of the community.
    • The definition of 50/50: 50/50 is the number of blocks rather than the target difficulty. The approach aims to balance both algorithms. It doesn't matter how much hash rate you have on one algorithm or another. The difficulty is adjusted so that the block is 50% one algorithm, and 50% the other. There are two approaches to doing the 50/50 split. The way it works in Grin is scaling the secondary algorithm off the difficulty of the first which is a more elegant way of keeping both in check. The existing difficulty adjustment algorithm remains in place which controls based on block time. That doesn't change, but a secondary check is needed to make a balance between the ratio of primary and secondary blocks. There are some elegant approaches like Grin uses, but the goal is not to make drastic block changes. The goal is to have 50% ETC hash and 50% Keccak-256.
    • The fairness of 50/50: 50/50 is fair because it would not be a permanent split, but a starting split. We will let the community decide what the right strategy is. Our preference will be to slide over to Keccak. The balance is trying to figure out the best interest of the community and the industry. In other words, as the example of Kadena shows, it aims to look after the ECT and make our ecosystem happy.
    • The rolling window: The blocks over the rolling window are trying to be maintained, and the difficulties are adjusted accordingly. In terms of blocks, the rolling window follows the difficulty adjustment algorithm window, so that both can be in sync. In a week, half the blocks are expected to be ETCHash and half the blocks to be Keccak-256. In that week, if the blocks are not 50/50, then the algorithms wait, and difficulty is adjusted accordingly. There will always be a primary proof of work, which is the ECT hash. And that difficulty is what everyone would access by default. But there would be a second scaling factor for the second algorithm proof of work.
    • Hash rate penetration: Henry agreed that there are more efficient GPUs coming despite ASIC. But it doesn't matter because it's just computing purely on power. Based on the operating costs, slides 13 shows how many rigs would be supported. It's projected that if the ETC price were to climb 8 times to 60, then it would support 33% of the network. Miners don't like to operate at a loss or zero profits. So, the hash rate penetration is way below that number.
    • GUP mining ETH hash: There is no issue with the GUP mining ETH hash, ETC hash, or Keccak-256. The only issues are that some ASIC can't mine Keccak. There will indeed be more FPGAs coming, but there is no new mining ecosystem here.
    • Hybrid vs. pure Keccak: hybrid emerges as a backup solution. The hybrid is a compromise approach to secure the chain. In terms of ramping Keccak hardware, GPUs and FPGAs are capable of mining Keccak as well. This is proven by Grin. SHA3 will consume way more GPUs.
    • The 1049 proposal: Mining is an open business and protocols are offering incentives. There are many ways for ETC to remain open to mining after transitioning to a hybrid proof of work. 1049 was a better approach because we can implement that now, whereas the other approaches were a lot harder. ETC needs to do something to enforce security and has a path forward on its own. Hybrid is going to take a little bit longer. One proposal is to do hard work for the Keccak altogether. Some changes have already been made to integrate Keccak into the codebase.
    • Power efficiency: Power efficiency doesn't matter too much in this context, because it all depends on the relative hash rate. In a saturated situation, being inefficient doesn't work. Efficiency matters a lot in bitcoin. In terms of cost for acquisition and operation, the ASIC provides a much better solution than GPUs. The 1049 proposal is probably one of the best ways to secure ETC's future.
    • The difference in competitiveness: There's a difference in competitiveness. The security model only really works if it's most profitable. Keccak improves interoperability and the ability to do a lot of other things.
    • The combination of FPGAs and ASICs: There are plenty of ASIC manufacturers delivering SHA3 capabilities. IB link and Goldshell are making SHA3 capable products already. There was a small network called miner for the Islamic community. Related documentation is unavailable, but growth can be seen in that network. The increase certainly indicates the combination of FPGAs and ASICs. FPGAs are what's on the small SHA3 networks at this point.
    • The advantages of SHA3: The splitting 50/50 means that the group/the blockchain/ETC is immune to any inter affairs activity. The seesaw approach decided that one side can dominate. Based on the premise that 15-25 GPUs equals one FPGA, that has the impact of neutralizing a lot of GPUs.
    • Hash rate consumption: how many it can consume is based on operating costs. It would consume more GPUs on the SHA3 side, than on the ASIC side because SHA3 Keccak uses more computing power. Henry is unsure about what is going to happen to the GPU hash rate. But he believed SHA3 will consume more of the hash rate and make it more difficult for the older GPU. By moving to higher GPUs and higher hardware, we negate the impact of excess population of the ETC hash and ASICs which are huge amounts of hash power.
    • Rebranding of Ethereum Classic: The new version of the website is currently already released for those that are interested in the discord channel. There's a link on the website.

Action Points

  1. Henry will Publish ECIP Hybrid PoW by February 18.

  2. Bob is going to be stepping into the champion role for the 1049 proposal.

  3. ETC PoW implementation is needed ahead of the ETH2.0 merge.

  4. Discussion is needed with the community on the pros and cons of taking the hybrid as a backup.

  5. Difficulty checking and clients for the full implementation by the end of 2022.

  6. Creation of a comprehensive document that provides links and records the opinion shared by everybody throughout time (consensus TRACKER ). 7. Announcement of the public release of the new website on this call next week on the 22nd.

  7. Collection of any final edits or suggestions for the website from the community within a week.