In the previous article, we explained the need for separating the currency creation and record keeping consensus to correctly align the incentives of all the players and create an unit of account for the crypto world.
Division of power is one of the key concepts for Meter’s consensus design. In the physical world, gold miners have always been separate from the bankers and accountants who kept records of financial transactions. Similarly, the proof-of-value consensus separates functions to act as a hybrid of proof-of-work and proof-of-stake.
Proof-of-work has three main functions in Meter: 1. Creating new currencies, 2. Generating true randomness and 3. Instrumenting the notion of time. It is completely permissionless; anyone can join the network and start creating the Meter currency. However, in the proof-of-work consensus we have created a special tweak that ensures the cost of production for Meter tokens is uncheatable and stays the same. This means each Meter will only be created when it makes sense economically. The proof-of-work miners are like the physical world gold miners; they invest more in mining when the price of Meter is high and reduce their spending — or shift the computation capacity to mine other cryptocurrencies — when the price is low. Their profit-maximizing behavior will naturally create discipline in Meter’s monetary policy.
Proof-of-stake is responsible for record keeping in Meter’s blockchain system. To prevent censorship, Meter allows a big pool of validators (participating nodes that can potentially keep the transaction records). At the beginning of each epoch, a group of validators will be randomly selected from the validator pool to form a committee. During the epoch, only the validators within the committee can create and vote on the blocks, following a BFT style consensus. This ensures high performance and instant finality, both of which are desirable for financial applications.
The proof-of-work miner and proof-of-stake validators maintain separate proof-of-work and proof-of-stake chains. The two chains only interact at end of every epoch, when they cross reference each other. This ensures it is impossible to go back in time and create new chains (called a long range attack).
Because of the changes in the economic and game design, the proof-of-value consensus is able combine the benefits of both proof-of-work and proof-of-stake while avoid the problems in both. As we know, in Bitcoin or other proof-of-work cryptocurrencies, the network hashing rate directly correlates to the price or the market cap of Bitcoin. In Meter, the hash rate only correlate to the increment of market cap.
Such a difference makes Meter greener than any traditional proof-of-work cryptocurrencies. Based on our estimation, to support an economy similar to the United States, the annual money spent in mining would be similar to the annual budget of U.S. Mint and U.S. Engraving and Printing. Proof-of-work is able to introduce true randomness and a notion of time. These are important to prevent censorship and attack, and extremely hard to obtain in proof-of-stake.
The combined proof-of-value consensus is therefore permissionless, green to mine, and extremely fast, while remaining highly decentralized. We have a target of 1,000 transactions per second in a pool of more than 1,000 validators on a single chain. There is nothing preventing Meter from further unlimited scaling through sharding and sidechains.