EVG101 Part 9: Scalability II
Welcome to Part 9 of Everest’s crypto and blockchain guide! In today’s article, we will elaborate on more details related to scaling solutions, most notably sharding and Proof of Work.
To go back to our home page: link
To go back to the previous article on scaling solutions: link
Table of Contents
- What is sharding?
- What is proof of stake?
- What is the difference between proof of work and proof of stake?
- How does proof of stake help to scale a blockchain?
TLDR;
1. What is sharding?
Sharding is an on-chain scaling solution which splits the Ethereum blockchain into partitions, and each partition node, or shard, only verifies new transaction data based on the data it holds. The global state of the entire blockchain is then simply the combined state of all the shards that the blockchain has been split into. Each unique Ethereum account will belong to one specific shard and without cross-shard communication protocols, accounts can only transact with other accounts in the same shard. This allows for throughput of transactions to increase as the workload is now split between different nodes, as opposed to the current situation where a single node, the main chain, does all the work.
In the usual Ethereum blockchain, transactions are organized into blocks and the blocks are chained together. However, sharding brings another layer of abstraction; a bunch of transactions are organized into transaction groups, and a bunch of transaction groups are organized into blocks. This is what a transaction group looks like:
Each transaction group belongs to a particular shard, identified by the shard ID, and each transaction is between two accounts belonging to that shard. By having shards, many parallel transactions can happen at the same time within shards to improve performance. The global state of the blockchain can still be accessed by looking at the states of all the shards.
However, in the case where an account wants to transact with an account from another shard, cross-shard communication protocol is required. The cross-shard communication utilizes two important facts:
- Receipts are generated for every transaction in a shard
- These receipts are stored in distributed shared memory so that it can be easily accessed by other shards, but not modified.
2. What is proof of stake?
Proof of stake is a consensus algorithm where participants in a network offer up something valuable as a deposit in exchange for an opportunity to reap some reward. This deposit is forfeited if the participants are found guilty of fraudulent behavior. Users are incentivized to behave honestly as they stand to lose their deposit otherwise; they have something ‘at stake’.
Although the Ethereum network currently uses a proof of work consensus algorithm, they are planning to switch to proof of stake in the near future. In the case of Ethereum, participants in a proof of stake network offer up ether tokens as a stake in exchange for a chance to be selected to validate Ethereum transactions and thus receive block rewards and transaction fees. There are various ways to implement proof of stake, but in general, the larger the stake of ether held (as a fraction of all available ether), the higher a miner’s chance of being selected to mine the next block and to receive the associated rewards.
3. What is the difference between proof of work and proof of stake?
Proof of work is dependent on computational power whereas proof of stake is dependent on the stake of token held. This fundamental difference affects how centralization is built up in a crypto network.
In Bitcoin where proof of work is used, miners can accumulate a large amount of power in the network by simply buying more and more computational power. Therefore with enough money, one can gain control of the Bitcoin network by purchasing enough computational power. However, for a proof of stake system, one cannot simply buy more influence in the network since there are a limited number of tokens. If there are no sellers willing to sell, then buyers cannot acquire more tokens. Therefore, your ability to increase your stake, and thus influence in the network, is dependent on others.
Furthermore, there is nothing at stake for miners in a proof of work network: if the value of Bitcoin collapses, miners can generate revenue from another source using their computational resources. However, in a proof of stake system, if users find out that there is an individual or entity which acquired a majority of token and thus effectively controls the network, they would be likely to dump their tokens to minimize their exposure to risk. This would cause the value of the token to drop, negatively affecting the entity holding the majority stake of tokens. Therefore, a proof of stake system is more averse to centralization.
4. How does proof of stake help to scale a blockchain?
Proof of stake can help to scale a blockchain by making sharding a viable scaling solution. Sharding proposes to split the Ethereum blockchain into smaller pieces, which can make small individual shards vulnerable to a 51% attack under a proof of work consensus since each individual shard is backed by less computational power. This problem is avoided with proof of stake as there is no mining involved. Proof of stake also allows Ethereum to randomly shuffle and assign a node to a shard to prevent attacking nodes from choosing the shard they want to attack.
Concluding Remarks
In this article, we explored other details on blockchain scaling solutions including sharding, proof of stake, and how proof of stake enables sharding. This is all the content we have for now, but we’ll be back with more very soon. If there are any topics you would like us to cover, or any lingering questions after reading our guides, please let us know in the comments section and we will try our best to reply and include the answers in updated versions of our guide. And if you found our guide useful, please leave some claps!
For more information on Everest, please visit our website.
