What are Proof-of-Work and Proof-of-Stake?

Proof-of-Work (PoW) is an algorithm that protects distributed systems from misuse (DoS attacks, spam, etc.). There are two key points:

• The need to perform a reasonably complicated and time-consuming task.
• The ability to check the result easily and quickly.

PoW tasks aren’t meant for humans. These can always be solved by a computer in a finite time, but require substantial computing power. Meanwhile, it requires minimal time and resources to check if the result is correct.

PoW is a mechanism of reaching consensus (a common opinion on the valid version of the blockchain). Bitcoin creator Satoshi Nakamoto took the Hashcash’s idea as the basis and added a mechanism of adaptive complexity, where N (the required number of zeros) depends on the cumulative hashing power of the network participants. In Bitcoin the hash function is SHA-256. Simply put, PoW allows a node to check if a miner (node adding a block to the chain) did the calculations. The process includes finding a hash of the header (the part of a block that contains information about the included transactions and a link to the previous block) that satisfies the current complexity requirements.

Bitcoin’s Proof-of-Work gave rise to an entire cryptocurrency mining industry. It also formed a commercial incentive for companies to develop specialized mining equipment, which led to an increase in total computing power. Power consumption became a significant concern.

By now there are several alternatives to PoW. The most widespread of those is Proof-of-Stake.

Proof-of-Stake (PoS) is an alternative consensus algorithm first seen in 2012 in PPCoin, know known as PeerCoin. The idea of PoS is to use stakes as a resource that determines which node gets the right to add a block.

Similarly to PoW, nodes in PoS are computing hashes to satisfy the requirements, but in this case, the complexity is individual for the node and depends on the amount of tokens on its balance. The node with a bigger balance is more likely to generate the next block.

This approach requires less physical resources such as electricity and leaves out the ethical question of wasted computation capacity. To attack a PoS network an adversary would need to acquire a large amount of tokens, which makes the attack economically unreasonable.

PoS incentivizes participants to amass funds, which may lead to centralization of power in the network. A small group with sufficient funds may be able to dictate rules to other participants. There is also the Nothing-at-Stake problem, that makes PoS systems appear unreliable to many cryptoenthusiasts. An adversary can try to create a longer chain with non-existent resources. Other participants may support this chain since it won’t cost them any real resource. With this alternative chain, an adversary can block certain transactions and execute the double-spending attack.

Currently, projects are experimenting with hybrid systems that take features from both PoS and PoW. Many cryptocurrencies use PoW while the coins are created through classical mining and then shift to the PoS stage after the emission is over.

Yet, while PoS-systems are easier to implement and just as reliable, major cryptocurrencies are unlikely to shift from PoW anytime soon.

Proof-of-Work and Proof-of-Stake are the most popular consensus algorithm, but they aren’t the only. Here are some of the well-known alternatives:

• Proof of Activity — a standard hybrid model combining PoW and PoS.
• Delegated Proof of Stake — an algorithm similar to PoS, where participants can delegate their tokens to validators in exchange for a share in the block reward. DPoS is used by BitShares, EOS, and Tezos;
• Proof of Burn — an algorithm that involves sending tokens to an address from which they can’t be spent. By burning tokens in such a way the user gets a lifetime right to mine. Blocks are added by randomly assigned lifetime miners.
• Proof of Capacity — an implementation of “megabytes as resources” idea. Mining implies sharing one’s disc storage space.
• Proof of Storage — a similar algorithm, where the shared acts as a cloud storage for the network participants.