The ambitious change is expected to cut energy consumption by a factor of 1,000.
In a few weeks, Ethereum is slated to undergo the most significant change in its seven-year history. Until now, the Ethereum blockchain has been secured using a method called "proof-of-work," which consumes more electricity than the entire nation of Belgium. Next month's switch to a new method called "proof-of-stake" is expected to cut Ethereum's energy consumption by a factor of 1,000.
The stakes are high. A botched transition could mean chaos for the many crypto projects built on top of Ethereum. A smooth transition would be the culmination of years of careful planning by Ethereum's core developers. Over the last year, developers have repeatedly pushed back the date of "the Merge" to give themselves more time to prepare. They completed a final dress rehearsal on August 10, clearing the way to make the switch in mid-September.
The most immediate consequence of a successful Merge will be to put the world's Ethereum miners out of work. Over the last seven years, thousands of people have bought high-end graphics cards to help maintain the Ethereum blockchain—and to earn newly created ether in the process. The new system for updating the Ethereum blockchain doesn't require the same kind of beefy hardware—or the massive electricity bill that goes with it. So the price of used graphics cards might continue to fall as Ethereum miners exit the industry.
But the switch to proof-of-stake is much more than just an energy-saving measure—it's a major overhaul of the Ethereum network. Ethereum founder Vitalik Buterin believes the Merge will lay the foundation for a series of future upgrades that will allow the network to handle a much larger volume of transactions in the coming years. But critics worry that the new scheme could cause the Ethereum network to become overly centralized—and hence vulnerable to government regulation.
From proof-of-work to proof-of-stake
At a high level of abstraction, here's how any blockchain works: Someone on the network proposes a block containing a list of recent transactions. Then other network participants verify that the block follows the network's rules. If a sufficient number of other network participants accept the block, it becomes the "official" next block in the chain. As long as most network participants are honest, users can have confidence that transactions accepted by a majority of the network won't be removed or modified later.
The big challenge for any blockchain project is preventing a malicious party from creating many sock puppet accounts to "stuff the ballot box," outvote the honest participants and thereby tamper with past transactions. Bitcoin's pseudonymous founder Satoshi Nakamoto's big insight—the one that made bitcoin possible—was that this problem could be solved using the principle of "one hash, one vote." On the bitcoin network, whoever has the most computing power—specifically, the capacity to compute SHA-256 hashes—has the most influence over which blocks get added to the blockchain. As long as honest miners have more hash power than malicious miners, users can be confident in the integrity of the blockchain—and hence in the integrity of payments made using the bitcoin network. (Check out our in-depth bitcoin explainer for details on how this works.)
When Vitalik Buterin launched Ethereum in 2015, he used a variant of Nakamoto's scheme. By that point, bitcoin mining was already dominated by specialized silicon optimized for computing huge numbers of SHA-256 hashes, locking ordinary bitcoiners out of the mining game. So Buterin developed a new mining algorithm designed to be "memory-hard"—and therefore difficult to accelerate with custom hardware. As a result, Ethereum mining is still largely performed using off-the-shelf graphics cards, allowing ordinary Ethereum users to participate.
But the economics of the two networks are fundamentally similar. As the values of bitcoin and ether have risen, it has become profitable for people to spend more and more money on mining hardware—and electricity—to generate new coins. While this has made the networks more secure, it has also meant that both networks consume astronomical amounts of electricity and hence drive more and more carbon emissions.
The bitcoin and Ethereum communities have responded to this issue very differently. Satoshi Nakamoto disappeared from public view in 2011. In his absence, bitcoin's culture has become increasingly conservative. Many bitcoiners adamantly oppose changing bitcoin's mining system, fearing that changes could open the door to centralization and ultimate government control. As a result, bitcoin is unlikely to move away from proof-of-work in the foreseeable future.
In contrast, the Ethereum community is still led by 28-year-old founder Vitalik Buterin, who has shepherded the network through a series of significant upgrades. Buterin has long recognized the environmental downsides of proof-of-work mining. Several years ago, he announced plans to transition Ethereum to proof-of-stake, which has been pioneered by several lesser-known cryptocurrencies.
While proof-of-work mining operates on the principle of "one hash, one vote," proof-of-stake is based on "one coin, one vote." Anyone who wants to participate in Ethereum's validation process must post ether as collateral, a process known as "staking." The more ether someone stakes, the more influence they have over which blocks get added to the Ethereum blockchain.
Every 12 seconds, a pseudorandom number generator selects a subset of stakers to form a committee to decide on the next block. One of them is designated to propose the next block, while the rest, called validators, verify that the new block follows all the rules of the Ethereum network. For example, if the block contains a payment transaction, the validators check that the source address has the required funds, that the transaction has the correct digital signatures, and so forth. If two-thirds of validators approve a block, it becomes part of the official blockchain.
Validators that faithfully follow these rules earn additional ether as a reward for their efforts, with the size of their reward proportional to the ether they've staked. On the other hand, if a validator tries to cheat—for example, by validating two different, incompatible blocks for the same blockchain "slot"—they will face financial penalties. If another validator posts evidence of such a cheating attempt, some of the cheater's collateral will be destroyed ("slashed," in Ethereum jargon), and the whistleblower will get a reward.
A difficult transition
While this scheme is pretty simple in principle, getting it to work right is fiendishly complicated in practice—especially because once the system goes live, many people will probe it for vulnerabilities they can exploit for personal benefit.
For example, it's crucial that the algorithm for random committee assignment produce truly unpredictable results. Otherwise, a malicious party might be able to manipulate it to gain a majority control over some committees. The validation algorithm also needs to deal flexibly with the normal turnover of validators without opening the door to either ballot-box-stuffing or denial-of-service attacks.
On top of that, the Ethereum team had to figure out how to make the switch with zero downtime—a feat some commentators have compared to replacing an airplane's engines while it's in the air. Much of the delay over the last two years has been due to evolving plans about how to handle the transition. Early plans called for proof-of-stake to be part of a larger "Ethereum 2.0" upgrade that would simultaneously change other aspects of the system. But the community has gradually scaled back its ambitions, deciding to implement proof-of-stake first and consider other enhancements down the road.
The plan the community ultimately settled on was to create a new proof-of-stake blockchain called the "Beacon Chain" that would initially operate independently of the existing Ethereum blockchain. The Beacon Chain launched in December 2020. Over the last 18 months, thousands of people and organizations have staked ether and have become validators on the new network. But they haven't had much to do since the Beacon Chain has no users. In a few weeks, the two networks will be merged together. At that point, the existing Ethereum network will be rechristened the "execution layer" of the new merged Ethereum network, while the Beacon Chain will become its "consensus layer," taking over the role currently played by proof-of-work miners.
The exact transition date isn't known; the Merge will happen when the network reaches a pre-set "terminal total difficulty"—a measure of the aggregate quantity of computation the miners have performed. Recent estimates suggest that will happen sometime between around September 15 and September 19, but the exact date depends on how much computing power miners bring to bear over the next month.
Fear of Uncle Sam
Last week, the Treasury Department announced it was slapping sanctions on Tornado Cash, an Ethereum "mixing service" that allows people to swap ether with anonymous strangers online. While advocates describe Tornado Cash as a way of protecting financial privacy, the US government considers it a tool for money laundering. The sanctions mean it's now a federal crime to use or interact with the Tornado Cash service.
Cryptocurrency advocates argue the Treasury Department has overreached because Tornado Cash is not a person or organization. Rather, it's a bit of code that's executed automatically by the Ethereum blockchain. Once the authors of Tornado Cash posted it to the blockchain, they lost the ability to change how it operates or shut it down. Advocates question whether the Treasury Department even has the authority to sanction a piece of software that's not under anyone's ownership or control.
A key question here is whether Ethereum miners and validators are affected by the sanctions. Above, I said that Tornado Cash is "executed automatically by the Ethereum blockchain," but that's not quite accurate. The code is executed by Ethereum miners. Individually, miners may lack the ability to change Tornado Cash's behavior, but their collective efforts make the Ethereum network—and hence Tornado Cash—function. Could the Treasury Department prosecute Ethereum miners who execute the Tornado Cash code as part of the mining process?
So far, the government has largely taken a hands-off approach toward bitcoin and Ethereum miners. This is largely a matter of practicality: A crackdown would probably push ether mining outside of the United States without fundamentally changing how the network operates.
The switch from proof-of-work miners to proof-of-stake validators could change things in a couple of important ways. Validators have to stake large sums of ether—the minimum is 32 ether, or more than $50,000. This requirement may mean that the types of people and organizations who become validators in the future will be different from the types of people and organizations who became miners in the past. In particular, bitcoin exchanges like Coinbase hold millions of dollars in ether and have been major participants in the Beacon Chain so far. These are large companies with strong ties to the US. They can't afford to thumb their noses at the US government.
Second, the new proof-of-stake system doesn't make it easy for a group of entities to suddenly drop out of the validation process the way miners can. If several parties drop out all at once, the network may interpret this as a denial-of-service attack and impose financial penalties on validators who go AWOL.
All of this means that the new group of Ethereum validators could prove more vulnerable to government coercion than the old Ethereum miners were. It's not hard to imagine the US government waiting until shortly after the Merge and then approaching a bunch of US-based validators and demanding that all of them refuse to validate transactions involving Tornado Cash addresses.
If a critical mass of validators complies, it could effectively prevent Tornado Cash from operating and set a precedent that the US government can block the operation of Ethereum smart contracts it doesn't like. That would create a kind of constitutional crisis for Ethereum because independence from government regulation has long been considered a core feature of networks like bitcoin and Ethereum.
To be clear, we don't know if the federal government will try something like this or how things would play out if it did. Coinbase CEO Brian Armstrong said he would shut down Coinbase's staking service rather than comply with demands to block specific Ethereum addresses. But the larger point is that the switch to proof-of-stake is far more than a technical tweak to improve energy efficiency. It's a fundamental change to how Ethereum is organized—one that could lead to big and unexpected shifts in power within the Ethereum ecosystem.
And this is precisely why the bitcoiners have resisted making major changes to their own network. The bitcoin community has a stronger ideological streak than the Ethereum community. More than anything, bitcoin's leaders care about preserving the network's independence from governments. And some see a risk that Ethereum's switch to proof-of-stake could undermine that autonomy.