The fifth delay of Ethereum’s difficulty bomb means that the finalization of POW may be delayed again

Wu Shuo Author | Wu Zhuocheng

Editor of this issue | Colin Wu

On August 20, Ethereum core developer Tim Beiko said in a review of the development team conference call that the difficulty bomb in December may be postponed, but there will be no new EIP before the merger. This is likely the fifth time the difficulty bomb has been postponed. Why is the difficulty bomb repeatedly postponed? What is the significance of its existence? What is its connection with the merger of ETH2.0?

1. The significance of the difficulty bomb

The difficulty bomb is a switch that causes the difficulty of mining to rise rapidly in Ethereum mining. The difficulty of Ethereum mining is not only related to the block time of the previous block, but also to the difficulty factor of the block. The calculation formula for the difficulty is listed below. The first half is the regular difficulty adjustment, which stabilizes the block time of each block at 13 seconds; the second half (the red frame part) is the difficulty factor, which determines the opening time of the difficulty bomb and the speed of its explosion.

Specifically, the difficulty factor includes the target block height for the difficulty bomb to be activated. When the block height does not reach the target value, the mining difficulty is only affected by the block time of the previous block; once the block height reaches the predetermined value, the block time will increase exponentially, and then the difficulty will be adjusted every 100,000 blocks mined.

On August 4, 2015, former Ethereum Chief Business Officer Stephan Tual first mentioned the difficulty bomb.

“A lot of people have been wondering how we will implement the switch from PoW to PoS during the Serenity phase. This will be handled through the newly introduced difficulty adjustment schedule, which will smoothly guarantee the hard fork point over the next 16 months… It will work like this: starting at block height 200,000 (around late August 2015), mining difficulty will begin to experience an exponential growth, and after about a year, the difficulty will increase significantly. At that time (around the time of the Serenity milestone), the large increase in mining difficulty will cause block times to increase.”

Why does the transition from PoW to PoS require a difficulty bomb? Understanding this issue will help us understand the order of the difficulty bomb and ETH2.0. First of all, we need to accept the fact that PoW will exit the stage of history after the completion of ETH2.0, and there will be no coexistence of the two consensuses. This has also been analyzed in Wu's previous articles, and the completion of the ETH2.0 merger phase is the end of the PoW mining method.

In this regard, there is a voice in the huge miner community that attempts to implement a fork after the merger of Ethereum 1.0 and 2.0. This will undoubtedly be a lose-lose situation for the entire Ethereum community, so the development team needs to find a way to prevent such incidents from happening. We know that due to the existence of Vitalik, the Ethereum development team is relatively easy to unite; but the miner team is not the same. Decentralization makes it impossible for all miners to form a community of interests. In other words, as long as miners can distrust each other, they will not have enough computing power to achieve a 51% attack.

The difficulty bomb is a conspiracy to make miners distrustful. According to Stephan Tual's estimate, it will take about a year for Ethereum to reach a level where it is almost impossible to mine blocks after the difficulty bomb is activated (the actual time may be shorter than this, and the time after each difficulty bomb is activated will be detailed later). This means that there will be at least a few months from the time the difficulty bomb is activated to the time when miners are unable to make ends meet, and this period of time is the time to disintegrate the miners' camp.

Considering that the higher the block height, the more difficult it is to achieve a fork, miners cannot wait until there is no output before starting the fork, but should start taking action as soon as the difficulty bomb is activated. However, imagine that as a rational individual miner, would you try to fork the main chain when the difficulty bomb is activated? In fact, you will worry: if some miners do not work together to fork, but continue to mine on the original longest chain, then their output will increase significantly, even exceeding the loss caused by the difficulty bomb. Once the fork fails, these "betraying" miners will make a lot of money, while the "united" miners will get nothing. Therefore, rational miners will choose to continue mining along the main chain (or choose to leave at a reasonable exit price). Even if they know that the final result is zero output, no one is willing to take the lead in taking risks.

By understanding the meaning of the difficulty bomb, we can naturally understand the order of its activation time and the ETH2.0 merger time. As mentioned above, the existence of the difficulty bomb is to prevent the collective fork of miners when Ethereum 1.0 and 2.0 merge. Therefore, if Ethereum has successfully completed the merger, then the difficulty bomb has no meaning. The reason why the final activation time of the difficulty bomb is regarded as a leading indicator of the end of PoW is based on this logic.

After the difficulty bomb is activated, the Ethereum merger must be completed within two or three months. If it cannot be completed, the team can only choose to suspend and postpone it. This has happened three times in history.

2 EIPs that have delayed the difficulty bomb

Historically, the difficulty bomb has been activated three times and postponed four times. The four postponements occurred in the Byzantium upgrade, Constantinople upgrade, Muir Glacier upgrade and the just-concluded London upgrade.

The above picture shows the changes in Ethereum block time in history. It can be seen that under normal circumstances, the block time is maintained at 13-14 seconds, but there have been three sharp increases, and these three times are when the difficulty bomb is turned on. After each turn on, the block time will show a sawtooth rise, which is because the difficulty will be adjusted once every 100,000 blocks are mined. Taking the initial 13 seconds as an example, 100,000 blocks took about 15 days, and the last time the time increased to 25 seconds. At this time, 100,000 blocks took nearly a month, so the sawtooth showed a trend of gradually widening. The three difficulty bombs were eventually suspended and postponed because the Ethereum development team was not ready to fully switch to PoS, and the increase in block difficulty under the PoW mode will lead to an increase in transaction waiting time and the probability of uncle blocks, thereby reducing the practicality and security of Ethereum.

The first difficulty bomb delay

In March 2017, the block height reached 3.7 million, the difficulty bomb was activated for the first time, and the block time increased rapidly. During this period, the difficulty was adjusted 6 times until it was extended to 30 seconds in October. At this time, if the difficulty bomb is not stopped as soon as possible, it will seriously affect the Ethereum ecosystem. After discussion by the development team, Ethereum launched a hard fork at a block height of 4.37 million - the Byzantine upgrade. The upgraded version includes the EIP-649 proposal, which delays the start time of the difficulty bomb by subtracting 3 million from the current block height as the pseudo block height used in the formula.

The real block height at this time is 4.37 million, but the pseudo block height used to activate the difficulty bomb is 1.37 million (437-300). A simple calculation shows that when the pseudo block height reaches 3.7 million again, that is, after 2.33 million (370-137) blocks - that is, the real block height reaches 6.7 million (437+233), the difficulty bomb will be activated again.

Second Delay of Difficulty Bomb

On January 15, 2019, Afri Schodeon, the coordinator of the Ethereum Constantinople upgrade, discovered that Ethereum's difficulty bomb had been activated as scheduled at a block height of 6.7 million, and "exploded" at a block height of about 7 million. The next day, Eric Conner, a core developer of Ethereum (the developer who launched EIP-1559), pointed out on Twitter that after the difficulty bomb exploded, the average block time had increased from 14 seconds to 15.5 seconds, and would accelerate.

In mid-February, the Ethereum block time rose to 20 seconds, and the problem once again faced Ethereum developers. The only way was to fork and upgrade to deal with the impact of the difficulty bomb. On February 28, the Constantinople upgrade completed the fork at a block height of 7.28 million, and the difficulty bomb crisis was once again temporarily resolved. The EIP-1234 included in the Constantinople upgrade not only postponed the difficulty bomb, but also an important change was to reduce the block reward to 2 ETH. The method of delaying the explosion this time is the same as the Byzantium upgrade, that is, simply reducing the pseudo-block height in the difficulty factor by 5 million. After the adjustment, the real height is 7.28 million and the pseudo-height is 2.28 million (728-500). When the pseudo-height returns to 3.7 million, that is, the real height reaches 8.7 million (370-228+728), the difficulty bomb will be activated again. This time, only 1.42 million blocks are needed, so the time is shorter than the previous time.

The third delay of the difficulty bomb

On October 5, 2019, the difficulty bomb was triggered early at block height 8.6 million, and in mid-December, the block time rose to 17 seconds. Ethereum developers had to hard fork again at height 9.2 million shortly after the Istanbul upgrade. This is the Muir Glacier upgrade, which has only one improvement proposal (named after the glacier because the upgrade only targets the difficulty bomb, and after the bomb fully explodes, the block reward will be "frozen", and the fifth postponement considered recently may also be named in this way), namely EIP-2384, which aims to postpone the difficulty bomb by another 4 million blocks, that is, 13.2 million height, about 611 days.

There is another sentence in the proposal that attracts attention: it is best to postpone the difficulty bomb again until the expected release date of the final version of the ETH2.0 tool (finality gadget).

The fourth delay of the difficulty bomb

This time, the Ethereum development team did not wait for the difficulty bomb to explode. On August 5, 2021, Ethereum launched the London upgrade at a height of 12.965 million, and the difficulty bomb was postponed again. This upgrade includes EIP-3554, which reduces the pseudo-block height in the difficulty factor by 9.7 million, or 2.77 million, from about 12.47 million (this height is not the current actual height, but an estimated height set when the protocol was proposed). Therefore, the difficulty bomb will be activated again after about 930,000 (370-277) blocks, which takes about 4.7 months (officially estimated to be launched in the first week of December) based on a block time of 13 seconds.

The fifth time EIP is proposed to postpone the difficulty bomb?

In July, Ethereum core developer Tim Beiko said: Assuming we have a non-merge upgrade (author’s note: assuming that the Shanghai upgrade in December cannot complete the 2.0 merge, in fact this assumption is likely to happen), we need to decide whether to include anything other than another difficulty bomb delay (author’s note: if other upgrade protocols are included, then keep the name "Shanghai", otherwise consider using the "Glacier Theme" naming, for the same reason as the Muir Glacier upgrade).

On August 20, Tim Beiko said in a review of the Ethereum core developer conference call that basically all teams agreed that in addition to the consensus changes for the merger of Eth1 and Eth2, there are other important things that need to be done in the next few months, most of which are client performance optimizations for the merger, better separation of consensus engines, etc. In view of this, the participants agreed not to perform a feature fork in December, which means that the difficulty bomb may be postponed and other small changes (one line) will be added, but no new EIP will be generated before the merger.

Now it seems that the difficulty bomb will be postponed again, which is consistent with our previous analysis of its connection with the Ethereum merger. In short, the merger means the end of PoW, and the complete explosion of the difficulty bomb means the completion of the merger. In previous articles, we analyzed that the end of PoW will not be until the second quarter of 2022 at the earliest, and the difficulty bomb will only take two or three months from the start to the complete explosion. Therefore, if the difficulty bomb is activated in December, Ethereum must complete the merger in the first quarter of next year, which is a difficult task.

The last question is, will the Ethereum development team propose a new EIP to postpone the difficulty bomb? In the past four times, developers have done this by reducing the pseudo-block height in the difficulty factor, and modifying the code requires proposing an EIP. How to postpone the difficulty bomb without proposing an EIP requires further explanation from the development team (the author speculates that Tim Beiko means that the December upgrade only includes the EIP to postpone the difficulty bomb and no other new proposals, so consider changing "Shanghai" to other "Glacier Themes").

Image from U.today

Welcome to read Wu's selected reports: Huobi exclusive report, Binance exclusive report, Bitmain series, supervision and card freezing series, Filecoin series, currency circle chaos exposure, mining farm supervision dynamics, etc.

Risk Warning

According to the "Risk Warning on Preventing Illegal Fund Raising in the Name of "Virtual Currency" and "Blockchain"" issued by the China Banking and Insurance Regulatory Commission and other five departments, readers are requested to abide by the laws and regulations of their region. The content of this article does not endorse the promotion of any business or investment activities. Investors are requested to raise their awareness of risk prevention. Wu said that the content of the blockchain is not allowed to be reproduced or copied without permission, and violators will be held accountable.