What Is Proof of Stake?
In 2020, the first phase of Ethereum 2.0 will go live, marking an overhaul of the existing Ethereum 1.0 blockchain and notable improvements in scalability and accessibility. The core of the Ethereum 2.0 architecture is the Proof of Stake (PoS) consensus mechanism, which will replace the existing Proof of Work (PoW) consensus mechanism.
PoS comes with a number of improvements, including: energy efficiency, lower barriers to entry, stronger crypto-economic incentives, and greater revenue-generating capabilities for a broader set of users. This article aims to clarify what Proof of Stake is, how it will be implemented in Ethereum 2.0, and how ETH holders can anticipate interacting with the new architecture.
Explore our Ethereum 2.0 Knowledge Base for Eth2 FAQs, resources, and more.
Understanding Consensus Mechanisms
In distributed systems, a consensus mechanism is the method by which the network agrees on a single source of truth. Unlike in centralized systems, where a source of truth is decided upon by a single controlling entity, distributed systems rely on large numbers of autonomous authorities to cooperate in the maintenance of a single network. These distinct nodes must have a computational mechanism by which to arrive at an agreement of what the most recent and accurate record of data is. To drive the point home, these distributed networks must all adopt an identical cryptographic mechanism to arrive at consensus.
Proof of Work Consensus Mechanism
The Proof of Work (PoW) consensus mechanism is currently the most widely-used consensus mechanism and arguably the best understood. Pioneered by Satoshi Nakamoto with the release of Bitcoin in 2008, PoW has so far powered the majority of highest-profile blockchains, including Ethereum.
For an emerging technology like blockchain, PoW has proven an extremely secure and trustworthy consensus mechanism. The basic components of PoW are miners and energy. Miners are the individuals or entities that maintain the network by running and managing nodes (computers). Miners direct nodes to expend electricity in the form of computational energy to solve increasingly complex mathematical problems. The miner that solves the problem first earns the right to add a block of transactions to the ever-growing chain of consecutive blocks, creating a single and verifiable history of data on a PoW blockchain.
The expenditure of computational power costs money in the form of electricity––on top of the initial hardware costs of setting up a functional node. The cost of being a miner, however, is made worthwhile by block rewards. When a miner successfully mines a block into existence, they receive a block reward in the form of the blockchain’s native coin (i.e. BTC, ETH, etc.).
As more miners begin to run nodes on a blockchain, the hash rate (i.e. computing power of the network) increases, meaning the next block may be mined into existence a little faster than the previous. The network attempts to maintain a consistent block time (the time between each block); Ethereum is mined every ~14 seconds and Bitcoin is mined every ~10 minutes. If the hashrate increases and a block is mined a bit faster than before, therefore, the difficulty is increased automatically for the next block, ensuring it becomes a bit harder to mine a block and the block time is re-stabilized. The difficulty regularly adjusts after every block so the block times stay relatively stable.
As miners leave the network (which can happen for a variety of reasons, including a decrease in the USD value of the native coin), the difficulty will decrease, meaning it becomes easier for miners to mine blocks and receive rewards. This decreased difficulty serves as an incentive for more miners to return to the network, ensuring the network remains strong and sufficiently decentralized.
PoW blockchains have proven extremely resilient and secure. The incentive against a malicious actor attempting to compromise a PoW blockchain is the cost of electricity required to generate the sufficient amount of computational energy to take over a majority hash rate. The combined computational power required for an individual to compromise a well-established PoW blockchain like Bitcoin or Ethereum would cost an extraordinary amount of money, and may not even exist.
Though simple and secure, PoW chains face three main challenges: accessibility, centralization, and scalability.
Accessibility: The barriers to entry to becoming a PoW miner are high. Proof of Work chains require a substantial amount of energy to maintain. A miner must purchase, set up, and maintain all the necessary hardware to run a PoW mining rig. Additionally, PoW mining is extremely energy-intensive. Not only is the underlying mechanism inefficient from an energy standpoint, but it further increases the barrier to entry. To earn significant block rewards, it is better for a miner to live in a region with lower electricity costs. Additionally, jurisdictions often offer lower electricity costs to corporations, meaning a miner who wishes to maximize their profits would need to form a company and purchase enough mining hardware to offset the effort and associated costs. Altogether, energy inefficiency, variable electricity costs, hardware costs, and corporate electricity breaks all present significant barriers to entry for most would-be miners.
Centralization: Barriers to entry for mining can have the adverse secondary effect of greater centralization of miners. As it gets more costly and less profitable to become a miner, the network naturally sees a concentration of mining into two categories. First, large mining conglomerates that operate in areas with low electricity costs and cold weather (to reduce the cost of manually cooling mining hardware) such as Mongolia and Siberia. Second, mining power is centralized in the hands of mining pools. As it becomes less profitable for most people to mine individually, they buy hash power from a mining pool, which operates as a single mining entity. By the end of 2019, over 50% of blocks on Ethereum were mined by just two mining pools.
Scalability: In the current Ethereum Proof of Work chain, each block is mined consecutively. Each block can only contain a certain amount of data, known as the block size. This means that if there are more pending transactions than can fit into a block, the transactions that do not make it into the next block to be mined must “wait” for the following block for another chance to be included. On Ethereum, a block is mined once every ~14 seconds, but during particularly high transaction events, some users could wait hours for their transactions to be processed.
Finding Solutions with Proof of Stake
Proof of Stake is a different kind of consensus mechanism blockchains can use to agree upon a single true record of data history. Whereas in PoW miners expend energy (electricity) to mine blocks into existence, in PoS validators commit stake to attest (or ‘validate’) blocks into existence.
Validators are the participants on the network who run nodes (called validator nodes) to propose and attest blocks on a PoS blockchain. They do so by staking crypto (in the case of Ethereum 2.0, ETH) on the network and make themselves available to be randomly selected to propose a block. Other validators then “attest” that they have seen the block. When a sufficient number of attestations for the block has been collected, the block is added to the blockchain. Validators receive rewards both for successfully proposing blocks (just as they do in PoW) and for making attestations about blocks that they have seen.
The crypto-economic incentives for PoS are designed to create more compelling rewards for proper behavior and more severe penalties for malicious behavior. The core crypto-economic incentive boils down to the requirement that validators stake their own crypto––i.e. money––on the network. Instead of considering the secondary cost of electricity to run a PoW node, validators on PoS chains are forced to directly deposit a significant monetary amount onto the network.
Validators accrue rewards for making blocks and attestations when it is their turn to do so. They are penalized for not following through with their responsibilities when it is their turn to do so – i.e. if they are offline. Penalties for being offline are relatively mild and equate to about the same as the expected rewards over time. So, if a validator is participating correctly more than half the time then her rewards will be net positive.
Should a validator attempt to attack or compromise the blockchain by trying to propose a new set of data history, however, a different penalty mechanism kicks in: a substantial portion of their staked amount will be slashed (possibly up to the whole amount of stake) and they will be ejected from the network. The result is a tremendous financial risk of a failed attack by a validator. To draw an analogy to PoW, it would be as if a miner who failed an attack on a PoW chain was forced to burn down her entire mining rig instead of just eating the cost of the electricity she spent on a failed attack. Furthermore, this architecture places the security of the network directly in the hands of those maintaining the network and holding its native crypto-asset in the protocol itself..
Proof of Stake addresses the three issues of PoW chains discussed earlier – accessibility, centralization, and especially scalability:
Accessibility: Proof of Stake blockchains do not require validators to worry about the initial hardware costs or pay attention to electricity rates in the same way miners on PoW chains must. It is, therefore, a significantly lower barrier to entry for an individual to run a validator node on a PoS chain than run a mining node on a PoW chain. There is, however, a notable barrier to accessible entry for PoS. Validators must stake a minimum amount of crypto to run a full validator node. For Ethereum 2.0, for example, this amount is 32 ETH ($6,500 at the time of writing). For many, that is a significant amount of money and a deterrent to active participation. In the same way PoW chains have mining pools, however, there will be staking pools that aggregate the funds of participants unable or unwilling to stake 32 ETH. The pool will stake on their behalf, and they will receive rewards as a percentage of their stake.
Centralization: With reduced barriers to entry and the elimination of concerns about minimizing electricity costs, PoS networks are significantly more decentralized at the node level than PoW networks. Participation in a PoS chain requires only a non-zero amount of crypto, an internet connection, and a computer (or phone/tablet). That opens up the doors of participation and revenue generation to a much larger group of people. Additionally, economies of scale are far lower in PoS economics than PoW. In PoW systems, the more hash power a miner controlled, the greater the % of rewards he would be able to receive. In PoS, a validator’s % return stays constant whether she manages 1 node or 1,000.
Scalability: Proof of Stake alone does not improve scalability. However, PoS architectures allow the implementation of a scalability solution known as sharding without reducing security. Sharding is a database scaling mechanism in which a blockchain is partitioned into multiple shard chains, each of which is capable of processing blocks. This relieves the blockchain from having to process each block simultaneously, and instead allows multiple blocks (and, in other words, more sets of data) to be processed all at once. With Ethereum 2.0, for example, sharding will partition the blockchain into 64 separate shard chains, meaning the network will process transactions at minimum 64x the transaction speed of the original PoW chain.
In PoW networks, sharding would help scalability, but would have a consequential impact on the security of the network. Dividing a PoW network into shard chains means each chain would require less hash power to compromise. PoS chains, however, “know” who the validators on the network are (more specifically, there is an address attached to each deposit, and therefore to each validator node). Through a provably random algorithm, therefore, PoS chains can ensure that the validators chosen to validate blocks on different chains are random, effectively statistically eliminating the chance that one validator controls enough stake to compromise the data in a block. Whereas PoW requires the tradeoff of security to achieve scalability, PoS networks can achieve both through sharding.
Proof of Stake on Ethereum 2.0
Ethereum 2.0 is a Proof of Stake chain that will go live in phases, starting with Phase 0 in 2020. Phase 0 of Ethereum 2.0 will launch what is called the beacon chain, which will establish and maintain the Proof of Stake consensus mechanism.
In Ethereum 2.0, the PoS consensus mechanism will require validators to stake 32 ETH to run a validator node on the network. Each time a block is set to be proposed, at least 4 and up to 64 random committees of 128 validator nodes will be selected from the entire pool of validators to attest the block. (This is provably secure, and there is less than a 1 in a trillion chance that an attacker controlling 1/3 of the validators on the network would control ⅔ of the validators in a committee to successfully execute an attack).
In order to become a validator on Ethereum 2.0, validators will deposit 32 ETH into the official Ethereum 2.0 deposit contract, which has been developed and released by the Ethereum Foundation. Validators will need to stake 32 ETH for each validator node they wish to run.
In Phase 0 of Ethereum 2.0, rewards for proposing and attesting will not be distributed to validators until the minimum threshold of staked ETH and committed validators is reached to launch the network. The network will require at least 524,288 ETH to be staked, divided among at least 16,384 validator nodes. Once the threshold is live and the genesis block is created, rewards will begin to be distributed to validators.
Read the Ethereum 2.0 Staking Report
With thanks to Mally Anderson, Ben Edgington, and James Beck.