ParallelChain is launching a new layer-one public blockchain boasting a TPS of 80,000. This makes it a top candidate for developers wishing to build on public chains or those who want to seize on the unique advantage of developing apps that work across ParallelChain’s public and permissioned blockchain networks.
To help achieve these speeds, ParallelChain centralizes some consensus-making. However, it improves on this approach by regulating centralization to stay within certain limits, and by making the privileged parties democratically accountable.
Tokenomic incentives play a very substantial role in delivering this outcome. They are used to keep voting power in the system well distributed and balanced according to ParallelChain’s intended design, without resorting to direct protocol controls on agent behavior, or voting power allocation.
This piece highlights some of the key tokenomic mechanics involved.
Consensus with 3 validator classes
ParallelChain’s consensus process has three tiers of seniority.
The most senior class — the Governing nodes — are allowed to exercise a greater share of voting power, among a small set of nodes. The most junior class — the Beta nodes — are allowed to exercise a smaller share of voting power but are much more numerous. In between them, in terms of both the number and expected voting power of each node, sit the Alpha nodes.
This design concentrates a greater share of consensus-making among the smaller sets of senior validators. That is what helps to achieve the higher transaction throughput on the network. Unlike other centralized chains, however, ParallelChain then adds democratically accountable checks to these privileges using a mix of Know Your Customer (KYC), protocol-controlled governance and economic incentive mechanisms:
- Governing and Alpha nodes are both subject to KYC requirements, and eventually pre-existing service requirements, too.
- Ascension to Governing and Alpha classes will be subject to a protocol-controlled governance process open to all token-holders on the whole network, including Beta nodes.
- Beta nodes are large in number and not subject to many checks. That keeps this body open and democratic.
- Validation incentives are engineered to avoid any class growing too powerful, and keep the stake balanced so that, on average, any class can be ruled by a quorum of the other two.
- The incentive design also ensures no node becomes disproportionately powerful within its class.
Managing concentration using a “triple cap” incentive design
To keep classes balanced, we have to address the “concentration of stake problem” that faces blockchains using a delegated proof-of-stake (DPoS) consensus mechanism.
The problem is motivated by the wish to avoid concentrations of stake forming. That keeps consensus-making on the network decentralized in the long run. In ParallelChain’s case, we need to harness similar properties to avoid carefully assigned and balanced privileges from moving far outside their intended bounds.
The Cardano team addresses the concentration of stake problem by applying a cap to the share of validation rewards claimable by any individual node. The cap is set to the share of stake that would be held if the stake were equally distributed among an ideal number of nodes. So if this ideal node number is 100, the cap is set at 1/100, or 1% of all the stake on the network.
Under this scheme, if a node has attracted 2% of the stake, it can still only claim 1% of rewards. If another under-saturated node has 0.5% of the stake and increases its stake to 0.75%, this node will get an extra 0.25% from the rewards pool.
This design makes it very hard for node operators to sustain over-saturated states. Delegators will be looking for attractive deals elsewhere and competitive under-saturated nodes will be able to offer delegators a higher return, attracting them away from over-saturated nodes.
This builds an intuitive picture of the tendency this reward design creates. In the example given, it pushes the system towards a state running on 100 nodes, each controlling about 1% of the stake.
To make this work for ParallelChain’s 3-class validator system, we just have to choose caps that target the number of nodes reserved for each class, given an assignment of one-third of the voting power to each class in the target state.
So, if we have 12 Governing nodes, their caps are set at 33%/12, and if we have 100 Beta nodes, their caps are set at 33%/100.
These incentives act at the node level. That may not be sufficient to stop a class from getting out of balance as a whole. To mitigate this, additional disincentives activate on over-saturated nodes if and when their class is over-saturated.
Conclusion
This article introduces how ParallelChain has harnessed limited centralization and subjected it to democratically accountable checks. We saw how tokenomic incentives provide ways to guide the system to the desired state without the need for hard controls on the network.
But there are plenty more tokenomic features at work! Read more here.
Find ParallelChain on: